Linzumi (Linzumi, Inc.)

June 25, 2026

Public disclaimer: This is an independent public teardown based solely on publicly available information retrieved on 2026-06-25. It is research support, not investment, legal, tax, or financial advice. Company-provided statements are labeled as claims, not verified facts. Linzumi did not participate in or review this report. Errors and omissions are possible; final decisions remain with the reader.

The company trying to keep coding agents on your machine still has to answer who owns the leverage. Linzumi pitches a team chat where humans direct Codex today, Claude Code later, and local agents under permissioned, auditable guardrails. The founder signal is not fake: Sean Grove's OneGraph acquisition by Netlify is independently on the record, and AI coding adoption is already moving from curiosity to default workflow. But the product lives inside a stack controlled by model vendors, IDEs, and trust gates it does not own. The teardown tests whether that tension leaves room for a durable control plane or only a clean interface over someone else's agents. Read it before the first founder call turns timing into conviction.

What The Company Does

Linzumi presents itself as a team chat and control surface for directing AI coding agents that run on a user's own machine, with humans able to review, approve, and steer work in a shared thread (Linzumi homepage) company claim. The company says Codex support is live and Claude Code support is coming soon on its Y Combinator launch page (Y Combinator launch) company claim.

The buyer is implied to be an engineering team, founder-led product team, or security-conscious organization that wants visibility and control over local coding-agent work. Linzumi publishes a free Personal plan, a flat $100 per month Company plan, and custom Enterprise pricing (Linzumi homepage) company claim. The legal pages identify Linzumi, Inc. as the counterparty and describe GitHub permissions, data handling, AI processing, and subprocessors including OpenAI, Anthropic, Google, AWS, and Tailscale (Linzumi terms; Linzumi privacy policy; Linzumi subprocessors) company claim for operating practice and source-backed fact for what the public policies state.

The company is a Y Combinator Spring 2026 startup founded in 2025, with Sean Grove publicly named as founder and a YC-reported team size of three (Y Combinator profile) source-backed fact. Public evidence does not show named customers, revenue, retained usage, or the other two team members.

Key Takeaways

Investor reaction
Linzumi earns a first-call look because a repeat developer-tools founder is attacking a timely, painful coordination gap in AI coding, but the public record keeps every fund-returning input below founder quality and category timing unproven.
Current verdict
hold with medium confidence. The right action is one proof-gated founder call, not deep diligence and not a check.
Why investors might lean in
Sean Grove's OneGraph exit is independently corroborated (Netlify) source-backed fact, Gartner forecasts 75% of enterprise software engineers using AI code assistants by 2028 (Gartner) source-backed fact, and Stack Overflow reports broad developer AI-tool adoption while showing team collaboration remains weak (Stack Overflow) source-backed fact.
Why investors might pull back
There is no independent customer, revenue, usage, or retention proof; the $100 flat price sits on third-party model and infrastructure dependencies; Claude Code already markets Slack-based team entry points (Claude Code) source-backed fact; security and trust claims are company-controlled; and YC reports three team members while public company profiles show only one.
Highest-leverage fix
Bring one independently checkable retained paying-team cohort, segmented by whether teams chose Linzumi for local or vendor-neutral execution, paired with gross-margin math for the $100 plan.
Best next move
Take one time-boxed founder call focused on retention, gross margin, team coverage, and enforceable trust. If those answers remain narrative-only, move to pass and revisit only on the named proof package.

What Makes This Potentially Fundable

The fundable version of Linzumi is not "chat for Codex." It is the default team control plane for AI engineering work: the durable thread where humans and agents coordinate, the local runtime stays trusted, and accumulated team context becomes the switching-cost asset. If that layer compounds, Linzumi could sit next to developer-workflow and collaboration platforms rather than below them.

The category timing is real. Gartner's forecast and Stack Overflow's survey show AI coding tools moving into mainstream developer workflows, while Microsoft reports Copilot reaching 1.8 million paid subscribers in fiscal 2024 and more than 20 million users in fiscal 2025 (Microsoft 2024 annual report; Microsoft 2025 annual report) source-backed fact. Public software comps show the adjacent prize can be large: GitLab reported $759.2 million in fiscal 2025 revenue and Atlassian reported $5.215 billion in fiscal 2025 revenue (GitLab; Atlassian annual report) source-backed fact.

The honest gap is that none of those sources prove Linzumi's claim on the value. Public evidence does not show retained teams, willingness to pay, positive unit economics, a shipped context moat, or a named security owner. The current case is a strong founder and wave option, not an underwritten venture outcome.

The Four Holes To Close Before Fundraising

Hole	Investor Fear	What To Bring
No independent paying-team retention.	The upside is a story without proof of pull, and launch-week attention can vanish once the novelty fades.	Product analytics showing paying-team count, weekly agent-hours, and four-to-eight-week multi-human cohort retention.
Gross margin and metering are unknown at $100 per month.	A flat team price over third-party model and infrastructure dependencies could turn growth into a margin trap.	A unit-economics packet showing cost of goods sold, bring-your-own-key share, Linzumi-borne compute, metering, and expansion mechanics.
Platform vendors can absorb the adjacent surface.	OpenAI, Anthropic, Microsoft, GitHub, and IDE vendors own models, distribution, and developer workflow budgets.	A structural explanation for why vendor-neutral local trust and accumulated context stay outside single-vendor products, plus win-loss evidence.
Trust and team coverage are not visible.	The product asks buyers to trust agents operating on real machines and repositories, yet public trust artifacts and the security owner are missing.	Architecture, threat model, pen-test summary, SOC 2 status, and named engineers responsible for secure local execution.

Decision Snapshot

One-sentence company description: Linzumi is a Y Combinator Spring 2026 startup building a team control plane for directing local AI coding agents from a shared chat thread.
Screen: hold.
Confidence: medium.
Suggested next action: One tightly scoped founder call, with retention and unit economics requested before or during the call.
Why this matters now: AI coding tools are becoming mainstream while team-level agent coordination remains unsettled, so the category window is open but contested.
Investor-readiness diagnosis: Strong founder and timing, weak outside-the-room proof.
Best founder use of this report: Treat it as a proof-package checklist for converting a credible control-plane story into evidence investors can underwrite.

IC Disagreement Map

The evidence dossier was reviewed independently across five partner lenses: the power-law lens tested fund-returning upside, the prepared-mind lens tested category timing and product insight, the founder-jockey lens tested team fit and execution, the risk-reduction lens tested critical flaws and value capture, and the long-horizon lens tested portfolio fit and durability.

Power-law partner
pursuelow confidence
A repeat founder with a verified exit, attacking a near-universal developer adoption wave, is an option-like first-call candidate even though the company proof is thin.
Would flip on
Retained paying teams, positive margin, and evidence of context lock-in would move them to stronger pursue; a confirmed capped-margin ceiling or vendor absorption would move them to pass.
Prepared-mind partner
holdmedium confidence
The local, neutral, team-thread seam is conceptually real, but demand for that specific seam and the enforceable trust layer are unproven.
Would flip on
Paid retention among teams that chose local or vendor-neutral execution would move them toward pursue; solo-only usage or an equivalent incumbent feature would move them toward pass.
Founder-jockey partner
pursuemedium confidence
At beta stage, the strongest evidence is the jockey: Sean Grove has a verified developer-tools exit and a relevant public AI background, which clears the first-call bar.
Would flip on
A named security or distributed-systems owner plus verification of the founder's exact OpenAI and OneGraph scope would raise conviction; an unnameable team or UI-only product core would move them toward hold or pass.
Risk-reduction partner
holdmedium confidence
No fatal flaw is confirmed, but value capture and platform absorption remain unresolved while traction is effectively absent.
Would flip on
Sustainable unit economics and retained multi-human teams would move them toward pursue; negative margin, vendor-native replication, or hollow trust controls would move them to pass.
Long-horizon partner
holdlow confidence
The only candidate compounding asset is the coming context layer, and public evidence does not yet show retention, ownership room, or durable value capture.
Would flip on
Shipped context features tied to retention, positive gross margin, and verified ownership availability would move them toward pursue; vendor encroachment, one-builder risk, or negative margin would move them to pass.

After the committee's evidence-request round, no partner changed vote or confidence. The final split remained two pursue, three hold, and zero pass, with the pursue votes explicitly framed as first-call option value rather than check conviction.

Scenario Range

Scenario	What The Company Looks Like In 3-5 Years	Falsifiable Trigger To Watch	Earliest Evidence
Strikeout	Linzumi becomes a convenient front-end or acqui-hire while teams default back to vendor-native coding agents and IDE tools.	A model or IDE vendor ships local, multi-human team-thread orchestration that reproduces Linzumi's wedge, or Linzumi's own cohorts prove solo-only retention with poor margin.	The first cohort export shows one active human per account, agent-hours tracking model releases rather than team workflow depth, and no metering path at the flat price.
Base	Linzumi retains a niche of security-conscious or vendor-neutral teams, but the market stays narrow and value capture remains capped.	Multi-week team retention exists in a bounded segment, gross margin is positive but expansion mechanics are weak, and the context layer creates moderate rather than compounding switching costs.	A small retained cohort and positive unit economics appear, but customers describe Linzumi as useful coordination rather than an irreplaceable system of record.
Home run	Linzumi becomes the default team control plane for AI engineering work, with humans and agents coordinating in a durable thread that accumulates context, permission history, and audit trust.	Multi-human retained agent-hours grow over time, the context layer is live and tied to retention, teams cite local neutrality as the reason they stay, and trust controls are enforced rather than cosmetic.	Product analytics, reference calls, a security architecture review, and early context-retention curves converge on the same explanation for stickiness.

What Investors Will Test

What We Looked For	Current Read	Investor Implication	Founder Prep Priority
A falsifiable fund-returning path.	The control-plane path is plausible but still a company claim supported by category timing rather than Linzumi-specific traction.	Investors will test whether the thread becomes a system of record or remains a user interface.	Bring the retained-team cohort and show what customers would lose by switching away.
Outside-the-room proof.	Founder identity, YC batch, OneGraph acquisition, and market adoption are supported; backers beyond YC, security claims, customers, and retention are not.	The first call should validate proof, not repeat the public story.	Prepare corroboration for backers, customer usage, security controls, and team roster.
Category insight.	The local, neutral, multi-human seam is conceptually differentiated but not proven as a buying reason.	Investors will ask whether the seam is a market or a temporary gap before incumbents fill it.	Segment win-loss and retention by local execution and vendor-neutral intent.
Jockey fit.	Sean Grove's developer-tools history is strong, but coverage for secure local execution and multi-user infrastructure is not visible.	Founder pedigree earns the call; team coverage decides whether the call advances.	Name the full team, responsibilities, and security owner.
Value capture and incentives.	Linzumi creates potential team value but depends on third-party model and infrastructure providers while charging a flat price.	Investors will not fund adoption that destroys margin or strengthens vendors' leverage.	Bring cost, metering, and expansion math.

Claim Reconciliation: Inconsistencies Investors Will Catch

Claim Or Metric	Where It Appears	Conflicting / Unreconciled Versions	Why Investors Flag It	How To Reconcile
Team size.	Y Combinator profile LinkedIn company page Crunchbase	YC reports a team size of three; public company and database profiles show one listed employee.	The hardest build needs more than a single visible founder, especially for security and local runtime work.	Publish or share an org chart naming the other two team members, roles, employment status, and owners for security and product infrastructure.
Funding and backer roster.	Linzumi homepage Crunchbase	The homepage lists six backers; Crunchbase shows a roughly $500K pre-seed row with Y Combinator.	Backer quality and ownership planning cannot be underwritten from homepage logos.	Provide cap table, signed financing documents, and investor names with permission to reference.
Company positioning.	Linzumi homepage LinkedIn company page	The homepage positions team chat for coding agents; LinkedIn carried stale API-composition copy.	Positioning drift makes investors wonder whether the pivot is operationally buttoned up.	Update public profiles and explain the pivot from API composition to agent coordination.
Security status.	Linzumi homepage Linzumi terms	The homepage markets security controls and enterprise trust; public terms reserve security commitments to signed agreements, and no public trust center was found.	Trust is the product's core promise, so unsubstantiated security language becomes a diligence blocker.	Publish a trust page or share security architecture, SOC 2 status, data-processing agreement, and enforcement evidence.
Product integration status.	Y Combinator launch	Codex is described as working today, while Claude Code is described as coming soon.	Investors will separate production integrations from roadmap claims.	Bring a live integration matrix with dates, usage, and known limitations.

Diligence Findings: Issues To Fix Before You Raise

2 deal-breaker3 high2 medium

No independently verifiable retained paying-team cohort is visible.
deal-breaker
Commercial and KPI diligence.
Fix
Prepare product analytics and customer references before the first investor call.
Gross margin and metering are unknown for the $100 Company plan.
deal-breaker
Financial and unit-economics diligence.
Fix
Show cost of goods sold, usage limits, metering, and expansion mechanics.
Trust claims are not supported by public trust artifacts.
high
Security, legal, and enterprise procurement diligence.
Fix
Publish or share security architecture, SOC 2 status, data-processing agreement, and pen-test summary.
The public team roster does not reconcile with YC's team size.
high
Team and technical diligence.
Fix
Name the full team and identify the owner of secure local execution.
Platform absorption is a live risk.
high
Competitive diligence.
Fix
Explain why neutrality, local execution, and context accumulation cannot be copied by model or IDE vendors.
Homepage backer logos are not corroborated beyond YC in public sources.
medium
Financing and ownership diligence.
Fix
Bring cap table and investor confirmations.
The context moat is roadmap, not evidenced.
medium
Product and retention diligence.
Fix
Show shipped C3 functionality and retention tied to accumulated context.

Data Room Readiness Checklist

Procurement-readiness score: partial.
Rationale: Public legal pages exist, but the artifacts a buyer's gatekeepers would expect for agents operating on real machines - SOC 2 status, data-processing agreement, subprocessor explanation, security architecture, pen-test summary, and data-retention controls - are not publicly complete.

Data Room Area	What Investors Expect	Current Read	Status	Priority
Company overview & corporate	One-pager, incorporation, good standing, structure	Public legal pages name Linzumi, Inc.; corporate documents are private.	`partial`	High.
Financials	Monthly profit and loss, 18 to 24 months of burn, three-year model with assumptions	No public financials or revenue evidence.	`missing`	Critical.
Cap table & funding history	Clean cap table, prior rounds, SAFEs or notes, 409A	Crunchbase lists a roughly $500K pre-seed row; homepage backers remain company-stated.	`partial`	Critical.
Legal & IP	Bylaws, board consents, IP assignments, material contracts	Public terms, privacy policy, AI terms, acceptable-use policy, and subprocessors are visible.	`partial`	High.
Product & technology	Roadmap, architecture overview, security and compliance docs	Homepage and launch pages describe product and roadmap; architecture and security enforcement are private.	`partial`	Critical.
Team	Org chart, key employment or advisor agreements, vesting	YC reports team size three; only one public employee is visible.	`partial`	Critical.
Customers & traction	Retention cohorts, annual or monthly recurring revenue bridge, opportunity list, references	No public customers, revenue, or retention.	`missing`	Critical.
Market & competition	Market model, competitive landscape, pricing	Market and competitor evidence exists, but the Linzumi-specific market share and willingness-to-pay model are private.	`partial`	High.

First-Call Agenda For The Startup

Time	Topic	Founder Goal	Evidence To Bring
0-5 minutes	Confirm what is shipped today.	Separate Codex, Claude Code, C3, security controls, and enterprise features into live, beta, and roadmap buckets.	Live integration checklist and product demo.
5-15 minutes	Retention and active teams.	Prove that whole teams, not just solo engineers, use Linzumi repeatedly.	Product analytics export and two referenceable customers or design partners.
15-25 minutes	Unit economics and pricing.	Show that the flat $100 plan can create durable margin and expansion.	Cost model, metering plan, billing examples, and bring-your-own-key split.
25-35 minutes	Security and team ownership.	Prove the hard local-control-plane build has accountable owners and enforced controls.	Org chart, architecture, threat model, and security owner background.
35-45 minutes	Defensibility and next milestone.	Explain why model and IDE vendors will not absorb the wedge and what evidence will prove that within 12 to 18 months.	Win-loss notes, context-roadmap evidence, and milestone plan.

Business Model And Revenue Signals

Personal free plan, Company $100 per month, Enterprise custom.
Company-controlled pricing copy.high
SourceLinzumi homepage
Public pricing exists; included usage and margin do not.
Terms reserve fees, usage, and security obligations to service terms and signed agreements.
source-backed fact for public terms.Company-controlled legal page.high
SourceLinzumi terms
This matters because homepage pricing alone does not define economics.
Crunchbase lists a roughly $500K pre-seed round led by Y Combinator.
company claim due aggregator source.Startup database.medium
SourceCrunchbase
Funding amount and investors need founder validation.
No public revenue, customer, or retention signal was found.
Public evidence absence.high
SourceLinzumi homepage Y Combinator launch Crunchbase
Revenue quality cannot be assessed from public sources.

Revenue and profit are not publicly available. Any investor model should treat Linzumi revenue, gross margin, retention, customer concentration, and payback as unknown until a data-room export is supplied.

Revenue Quality Checklist

Before this can move from hold to pursue, Linzumi should be ready to show the raw data investors will rebuild: paying-team count, monthly recurring revenue, gross revenue retention, net revenue retention, churn, customer concentration, cohort retention, agent-hours per team, cost of goods sold, gross margin, customer acquisition cost, lifetime value, and payback. Those figures need to tie out across billing exports, product analytics, the financial model, and customer references.

Funding And Ownership Context

Item	Public Read	Evidence Label	Diligence Request
Total raised / rounds	Crunchbase lists a roughly $500K pre-seed row.	`company claim` because the source is an aggregator.	Cap table, financing documents, and confirmation of all investor names.
Instruments (equity / SAFE / notes)	Not public.	`unknown`.	Instrument type, terms, valuation cap or price, and outstanding notes or SAFEs.
Investors	The homepage lists YC, Matrix, SV Angel, Decibel, Pioneer Fund, and Axiom Partners; Crunchbase lists YC.	YC is corroborated; non-YC names are `company claim`.	Signed SAFEs, investor confirmations, and permission to reference.
Post-money valuation	Not public.	`unknown`.	Current valuation and ownership math.
Cap table / ownership	Not public.	`unknown`.	Founder ownership, option pool, YC ownership, and available allocation.
Burn & runway	Not public.	`unknown`.	Monthly burn, cash balance, runway, and hiring plan.
Next-round plan	Not public.	`unknown`.	Target round, use of proceeds, milestones, and follow-on availability.

Founder And Team

Sean Grove
Public founder and CEO; repeat developer-tools founder.
Y Combinator profile Sean Grove LinkedIn Sean Grove GitHub
Evidencesource-backed fact for founder identity; public profile details are self-reported.
Confidencehigh
OneGraph acquisition
Prior founder outcome relevant to developer tools and API composition.
Netlify Y Combinator - OneGraph
Evidencesource-backed fact.
Confidencehigh
Prior YC companies
YC lists Sean Grove on Zenbox, OneGraph, and Linzumi.
Y Combinator - Zenbox Y Combinator - OneGraph Y Combinator profile
Evidencesource-backed fact.
Confidencehigh
Team size
YC reports team size three; LinkedIn and Crunchbase show one public employee.
Y Combinator profile LinkedIn company page Crunchbase
Evidencesource-backed fact conflict.
Confidencemedium
Public product repository signal
The Linzumi GitHub organization shows a public fork rather than original product core.
Linzumi GitHub
Evidencesource-backed fact.
Confidencemedium

Founder Competency Coverage

Competency	Sean Grove	What The Evidence Is
Domain depth	`evidenced`	OneGraph acquisition and developer-tools history are public; OpenAI post-training and alignment scope is publicly represented but still needs exact scope validation.
Technical build capability	`evidenced`	Public GitHub history and prior developer-tools company support builder credibility, while Linzumi product code remains private.
Product	`evidenced`	Linzumi beta, YC launch, and prior OneGraph outcome support product-building ability.
GTM / sales	`claimed`	YC launch copy and founder social surfaces show founder-led launch, but enterprise sales history for Linzumi is not independently evidenced.
Leadership / hiring	`claimed`	Prior company outcome and YC team size support the possibility, but the current roster is not public.
Fundraising history	`evidenced`	YC participation across multiple companies and the Crunchbase pre-seed row support fundraising familiarity.
Prior founding outcomes	`evidenced`	OneGraph was acquired; Zenbox is inactive; Linzumi is ongoing.

Secure local agent runner with credentials, permissions, scoped tunnels, and audit logging.
high
Team coverage today
Sean has developer-tools credibility, but no named security or distributed-systems owner is public.
What would close it
Named owner, architecture, threat model, and security references.
Multi-human real-time team coordination.
high
Team coverage today
Team size three is reported, but public roster and role coverage are not visible.
What would close it
Full team roster, roles, and product ownership.
Multi-provider agent orchestration.
medium
Team coverage today
Codex is claimed live and Claude Code is claimed roadmap.
What would close it
Production integration checklist and usage logs.
Continuous Context Compiler and retained context.
high
Team coverage today
C3 is described as coming soon.
What would close it
Shipped slice and retention tied to accumulated context.
Enterprise trust and procurement.
high
Team coverage today
Public legal pages exist, but public procurement artifacts are incomplete.
What would close it
SOC 2 status, data-processing agreement, subprocessor detail, and pen-test summary.

Public Professional Footprint

These public footprint checks carry low decision weight by design: they corroborate or weaken the competency table and sharpen founder-call questions, but none of them alone changes the verdict.

Signal	Sean Grove	What The Public Record Shows
Technical artifacts (repo level)	`evidenced`	Sean's public GitHub profile and repositories show a long technical footprint, while the Linzumi organization itself shows only a forked public repository (Sean Grove GitHub; Linzumi GitHub).
Professional social content	`evidenced`	Public profile and social surfaces show developer-tooling and Linzumi launch activity, but repetition of launch claims is not independent validation (Sean Grove LinkedIn; Sean Grove X).
Education / credentials	`claimed`	LinkedIn profile data includes education and YC-related history, but degree or institutional completion was not independently cross-checked (Sean Grove LinkedIn).
Publications / patents / certifications	`absent`	No role-relevant public patent, publication, or certification record was surfaced in this research.

Net read: the founder signal is strong enough for a call, but the public footprint does not close the secure local-control-plane owner gap.

Founder-Market Fit Read

What is promising: Sean Grove has an independently verified developer-tools exit, repeated YC history, public technical artifacts, and public AI-related representation that makes him unusually credible for an early agent-orchestration company.
What is missing: The exact OpenAI scope and dates, the specific OneGraph architecture he personally owned, the other two team members, and the security or distributed-systems owner remain unverified from public sources.
What to prepare: A concise founder-market-fit proof pack with role verification, architecture ownership, team roster, and references who can speak to the hardest technical build.

Linzumi, Inc.
Operating legal entity named in public legal pages.
Linzumi terms Linzumi privacy policy
Evidencesource-backed fact for public legal-page statement.
Confidencehigh
Follow-upConfirm incorporation, good standing, IP ownership, and cap table.
OneGraph
Prior company co-founded by Sean Grove and acquired by Netlify.
Netlify Y Combinator - OneGraph
Evidencesource-backed fact.
Confidencehigh
Follow-upConfirm Sean's exact role and owned architecture.
Zenbox
Prior YC company founded by Sean Grove, now inactive on YC.
Y Combinator - Zenbox
Evidencesource-backed fact.
Confidencehigh
Follow-upAsk what was learned and whether any obligations remain.

Traction

Customer Status Table

Customer	Strongest Evidence	Customer-Controlled?	Status	Paid?	Seats	Renewal	Confidence

Y Combinator Launch, Product Hunt, Hacker News, Crunchbase, LinkedIn, X, Indeed, Glassdoor, and general public web checks showed no named customer, case study, review, or paying-team record. This absence is decision-relevant because the investment case turns on retained team usage, but it is not surprising for a fresh beta and should be resolved directly on a founder call.

Traction Signals

YC Spring 2026 company profile and launch post.
source-backed fact for profile and launch existence; launch narrative remains company-authored.high
SourceY Combinator profile Y Combinator launch
Paying-team retention and usage.
Launch post showed 22 upvotes around retrieval.
source-backed fact as a launch proxy.high
SourceY Combinator launch
Conversion from launch interest to retained use.
Company X account and LinkedIn company page exist with early-stage small followings.
source-backed fact as social proxies.medium
SourceLinzumi X LinkedIn company page
Engagement and conversion to customers.
Digg captured YC ecosystem amplification around launch.
source-backed fact for amplification, not customer proof.medium
SourceDigg
Independent customer or buyer engagement.
No public revenue, customer, or retention metric surfaced.
unknown.high
SourceLinzumi homepage Crunchbase
Billing and product analytics export.

Product Hunt and Hacker News were checked and showed no matching public product launch. These absences matter less than customer retention because launch-channel choice is optional, but they keep public demand proof thin.

Hiring And Org Momentum

YC-reported team size.
source-backed fact.YC lists a team size of three.
SourceY Combinator profile
Public expectation is a team, not a solo project.
Public employee count.
source-backed fact conflict.LinkedIn and Crunchbase show one public employee profile.
SourceLinkedIn company page Crunchbase
The missing public roster is a diligence item, not proof the team does not exist.

LinkedIn Jobs, Indeed, Glassdoor, and applicant-tracking-system searches were checked and showed no matching job listings. These absences carry limited weight for an early beta, but they do not help resolve the team-depth question.

Role-mix read: unknown, because no open roles or full roster are public.
Momentum read (growth / steady / contraction / unknown): unknown, because launch activity is visible but hiring and customer momentum are not.

Traction Quality Read

Traction Dimension	Current Status	Good Enough For First Call?	Needed For Deep Diligence
Launch attention	YC launch and light social amplification are visible.	Yes, as context only.	Conversion from launch attention to retained usage.
Paid customers	No public evidence.	No, except as a question.	Paying-team count, invoices, and references.
Retention	No public evidence.	No.	Four-to-eight-week cohort retention with active humans per team.
Revenue quality	No public revenue evidence.	No.	Monthly recurring revenue, gross margin, customer concentration, churn, and expansion.
Team adoption	No public multi-human usage evidence.	No.	Product analytics proving multiple humans in a team use the thread repeatedly.

Competitive Landscape

Segment	Examples	Customer Alternative	Pressure On Company
Adjacent team-orchestration products	AgentsRoom, OpenACP, Crewly, Codezal.	Use a specialized agent-orchestration interface or bridge existing chat tools to local agents.	These alternatives prove the workflow is being attacked, even if none matches the full Linzumi wedge.
Model and IDE incumbents	Claude Code, OpenAI Codex, GitHub Copilot, Cursor.	Stay inside the model vendor, IDE, or GitHub workflow already in budget.	Incumbents own distribution, model access, and enterprise trust surfaces.
Hosted agent substitutes	Devin, Devin Desktop.	Use hosted or desktop agent fleets rather than a separate team chat.	They can satisfy "watch agents work" without Linzumi's local-trust claim.
Open-source and developer-built stacks	OpenHands, Cline, Aider, vibe-kanban, Claude Squad.	Compose open-source tools and existing chat apps.	Free or cheap substitutes pressure willingness to pay.

Category Visibility Snapshot

Google / United Statesteam chat AI coding agents local machine

Captured page-one results (2026-06-25): Reddit, VS Code local agents, GitHub open-source lists, and video results appeared.
Was the company present?: No.

Google / United StatesAI coding agent orchestration team collaboration

Captured page-one results (2026-06-25): AgentsRoom, AI-agent essays, listicles, and community results appeared.
Was the company present?: No.

Google / United Statesbest AI coding agents team chat 2026

Captured page-one results (2026-06-25): Best-tool roundups and AI-coding-agent lists appeared.
Was the company present?: No.

Google / United StatesLinzumi coding agents team chat

Captured page-one results (2026-06-25): Linzumi's YC launch and homepage appeared at the top.
Was the company present?: Yes.

Visibility read (inference, low confidence): Linzumi has branded visibility but not category visibility. That is normal for a fresh YC launch, yet it reinforces that current demand is not visible outside the founder's own channels.

Pricing And Competitive Benchmark

Alternative	What It Offers	Public Price Signal	Price Vs. This Company	Evidence Label
Linzumi	Team chat for local coding agents.	Free Personal, $100 per month Company, Enterprise custom (Linzumi homepage).	Flat team plan.	`company claim`.
GitHub Copilot	AI coding assistant and cloud agents inside GitHub and IDE workflows.	Public Copilot tiers include per-user pricing (GitHub Copilot).	Incumbent per-seat benchmark rather than direct team-thread pricing.	`source-backed fact` for public page content.
Claude Code	Terminal and IDE coding agent with Slack entry points.	Public page describes product and plans (Claude Code).	Platform-owned agent surface rather than neutral team layer.	`source-backed fact` for public page content.
OpenACP	Self-hosted bridge from chat tools to local agents.	Open-source/self-hosted positioning (OpenACP).	Free or self-hosted alternatives pressure Linzumi's price.	`source-backed fact`.
AgentsRoom	Desktop/mobile orchestration for coding agents.	Public pricing was not the focus of this check; product overlap is the benchmark (AgentsRoom).	Feature alternative, not a clean price comp.	`source-backed fact`.

Price positioning read: Linzumi's $100 flat team price is attractive if multiple people and agents retain, but risky if heavy usage creates Linzumi-borne cost or if users can get enough value from free and incumbent tools.
Price claims to correct or substantiate: The company should reconcile homepage price with any usage-based limits in legal or billing terms and show the gross-margin math.

Competitive Wedge

The best wedge is the combination of local execution, vendor neutrality, multi-human visibility, and audit or permission trust. That is meaningfully different from a hosted agent and different from a single-vendor IDE assistant. What still looks copyable is the chat thread, task kickoff, and agent orchestration view. The durable wedge must therefore come from enforceable trust and accumulated team context, not from the interface alone.

Risks And Open Questions

Risk	Severity	Evidence	What To Ask
No retained paying-team proof.	`deal-breaker` if unresolved after a fair beta window.	No public customer, usage, revenue, or retention evidence was found.	"Show the cohort where multiple humans in the same team keep using Linzumi after four or more weeks."
Margin trap at the flat price.	`deal-breaker` if gross margin is thin or negative with no metering.	Price is public; costs are unknown (Linzumi homepage; Linzumi subprocessors).	"Who bears model and compute costs, and what happens at high agent-hours?"
Platform absorption.	high.	Incumbent product pages show adjacent movement by Claude Code, Copilot, Codex, and Cursor.	"What structural reason keeps single-vendor platforms from copying the thread?"
Trust and security gap.	high.	Public legal pages exist, but public trust artifacts are incomplete; terms reserve security obligations to signed agreements (Linzumi terms).	"Are controls technically enforced, and what is the SOC 2 status?"
Team coverage gap.	high.	YC says team size three, while public profiles show one employee.	"Who owns local runtime, security, and multi-user infrastructure?"

Pre-Mortem: The Most Likely Obituary

Cause of death (one sentence): Linzumi died because model and IDE vendors absorbed the coordination surface before Linzumi converted local trust and team context into a durable, margin-positive control plane.
The causal chain from today to shutdown, in five steps:
1. Early usage proved mostly solo and convenience-driven rather than multi-human team workflow.
2. Claude Code, Codex, Copilot, or Cursor extended their own team and local orchestration surfaces into the same workflow.
3. Linzumi's flat price left too little margin to out-invest incumbents, and C3 did not ship deeply enough to create switching costs.
4. Enterprise buyers asked for trust artifacts and security ownership the small team could not produce fast enough.
5. Usage decayed to vendor-native tools, and the company became a talent acquisition or small utility rather than a category winner.
The earliest observable warning sign: Retained accounts show one active human per team, with agent-hours moving with model releases rather than with team workflow depth.
The question that defuses this chain today: "Show the retained multi-human cohort and the gross-margin math that proves Linzumi is a team control plane, not a solo convenience wrapper."

Decision-Critical Unknowns

Unknown	Why It Is Decision-Critical	Best Evidence	Decision Effect
Paying-team retention.	It determines whether the company has pull beyond launch and founder reputation.	Product analytics export and referenceable paying teams.	Strong multi-human retention would move the case toward pursue; solo-only or no retention would move it toward pass.
Gross margin under $100 flat pricing.	It determines whether adoption creates value the company can capture.	Cost of goods sold, metering, overage, and bring-your-own-key split.	Positive durable margin would support the land-grab read; negative or uncapped margin would trigger pass.
Security enforcement.	It determines whether the trust-first wedge is real.	Architecture, threat model, pen test, and SOC 2 status.	Enforced controls would raise confidence; UI-only controls would trigger pass.
Team coverage.	It determines whether the hardest build rests on one person.	Org chart, roles, public profiles, references, and ownership of the security build.	A credible team would raise founder-jockey conviction; unnameable or uncovered security ownership would lower it.
C3 and switching costs.	It determines whether the long-horizon moat compounds.	Shipped C3, retention tied to accumulated context, and customer switching-cost evidence.	Shipped sticky context would support pursue; perpetual roadmap status would reinforce hold or pass.

Diligence Questions

First Call

Question	Why It Matters	Good Evidence
How many paying teams retain for four or more weeks, and how many active humans are in each team?	It distinguishes team control-plane pull from solo beta curiosity.	Product analytics export, invoices, and references.
What is gross margin at realistic agent-hours on the $100 plan?	It tests whether pricing is a land-grab or a leak.	Cost model, bring-your-own-key share, metering, and overage rules.
What stops model and IDE vendors from absorbing the workflow?	It tests the structural moat.	Customer win-loss showing local neutrality and context retention as buying reasons.
Who owns security and local runtime beyond Sean?	It tests team coverage for the hardest build.	Named team members, roles, and references.
Are trust controls enforced and is SOC 2 attested, in audit, or aspirational?	It tests whether the enterprise wedge is real.	Architecture, threat model, audit status, and pen-test summary.

Follow-Up

Question	Why It Matters	Good Evidence
Is C3 shipped, and does retention improve with accumulated context?	It tests the compounding moat.	Live C3 demo and retention curves tied to context depth.
Which agent runtimes are production today versus roadmap?	It prevents roadmap claims from being treated as shipped product.	Integration checklist, changelog, and usage logs.
Who invested beyond YC, and what ownership is available?	It determines financing credibility and fund-returning ownership.	Cap table, financing documents, and investor confirmations.
Who is the budget owner for the Company plan?	It tests go-to-market clarity.	Buyer interviews, invoices, and budget source.
Why are legal and trust materials not easier to discover from the marketing site?	It tests self-serve and enterprise hygiene.	Updated footer links, data-processing agreement, and trust center.

Kill Criteria

Kill Criterion	Evidence That Would Trigger It
Margin trap.	Cost of goods sold meets or exceeds the $100 plan under realistic agent-hours, with no metering or expansion path.
Full platform absorption.	A model or IDE vendor ships native local, multi-human team-thread orchestration that reproduces the wedge.
Hollow trust build.	Permissions, audit logs, or scoped access are UI-only, with no enforced control plane or remediation plan.
Bus-factor and no security owner.	The team cannot name a credible owner for the security and local-runtime build, and the product remains founder-only.
Solo-only or no retention.	Beta usage after a fair window is mostly solo or churned, with no retained paying teams.

Double-Down Criteria

Double-Down Criterion	Evidence That Would Justify More Diligence
Retained multi-human paying teams.	Cohorts show whole teams using Linzumi repeatedly, with growing agent-hours and referenceable buyers.
Durable gross margin.	Unit economics show positive margin at the flat price or a credible metering and expansion path.
Enforced trust layer.	Architecture, pen test, and SOC 2 progress show real security controls.
Shipped context moat.	C3 is live, and retention or expansion improves with accumulated context.
Power-law ownership.	Cap table and round terms leave meaningful ownership and follow-on access available.

Founder Action Plan

Timeframe	Action	Output
Before the next investor conversation	Export paying-team count, four-week cohort retention, active humans per team, and weekly agent-hours, segmented by local or vendor-neutral buying reason.	Retention proof packet.
Before the next investor conversation	Build a unit-economics one-pager for the $100 plan.	Gross-margin, metering, overage, and bring-your-own-key analysis.
Before the next investor conversation	Name the team and security owner.	Org chart, public profiles, and role ownership.
Near term	Publish or share trust artifacts.	Security architecture, SOC 2 status, subprocessor explanation, data-processing agreement, and pen-test summary.
Near term	Define the structural defensibility thesis.	Written explanation of neutrality, local execution, context accumulation, and why vendors will not cross the seam.
Near term	Ship or evidence C3 progress.	Live context feature and retention-tied-to-context data.
As available	Verify financing and founder scope.	Cap table, investor confirmations, OpenAI scope, and OneGraph architecture references.

Decision

Screen: hold.
Confidence: medium.
Rationale: Linzumi combines a strong repeat founder, credible category timing, and a non-obvious local/team wedge, but every proof point that would make it underwritable - retained paying teams, gross margin, enforceable trust, team coverage, context compounding, and ownership - is absent or founder-controlled in the public record.
What would move this to pursue: A retained multi-human paying-team cohort, positive durable margin at the flat price, a named security owner, enforced trust controls, and evidence that local neutrality or accumulated context drives retention.
What would move this to pass: Negative or uncapped margin, solo-only or no retention after a fair window, UI-only trust controls, no security owner, or native incumbent replication of the wedge.
Recommended next step: One time-boxed founder call focused on the proof package, with a written decision immediately after the call.
Founder preparation standard: Bring raw evidence, not slides: product analytics, unit economics, security architecture, team roster, cap table, and references.

How We Would Miss This One

The miss scenario: Linzumi could be the one that got away if the public record is simply too early, Sean Grove's repeat-founder insight identifies the right local and vendor-neutral seam, and retained teams soon prove the thread becomes their system of record before incumbents decide to cross it.
Flip conditions: Multi-human retained teams, positive gross margin, a shipped context layer, enforced trust controls, and a credible security owner would flip the case toward pursue.
Revisit trigger / date: Revisit by 2026-09-25, or sooner if the company can provide the retention, margin, trust, and team proof package.

Source Log

Linzumi homepage
linzumi.com
Product positioning, beta status, pricing, security messaging, and company-claimed backers.
Retrieved 2026-06-25Company-controlled page.high
Y Combinator profile - Linzumi
ycombinator.com
Founder identity, batch, founding year, location, and team size.
Retrieved 2026-06-25Accelerator company page.high
Y Combinator launch - Linzumi
ycombinator.com
Launch narrative, Codex and Claude Code status, and launch proxy.
Retrieved 2026-06-25Accelerator launch post.high
Linzumi terms of service
serve.linzumi.com
Legal entity, GitHub scope, fees, security language, and arbitration.
Retrieved 2026-06-25Company-controlled legal page.high
Linzumi privacy policy
serve.linzumi.com
Data handling, retention, AI processing, and privacy rights.
Retrieved 2026-06-25Company-controlled legal page.high
Linzumi subprocessors
serve.linzumi.com
Model-provider stack and infrastructure vendors.
Retrieved 2026-06-25Company-controlled legal page.high
Crunchbase - Linzumi
crunchbase.com
Funding row, founder, and employee-count context.
Retrieved 2026-06-25Startup database (Crunchbase).medium
LinkedIn company - Linzumi
linkedin.com
Public employee count, follower count, and stale positioning.
Retrieved 2026-06-25Professional network company profile.medium
LinkedIn - Sean Grove
linkedin.com
Founder public profile and avatar source.
Retrieved 2026-06-25Professional network profile.medium
X - Linzumi
x.com
Company social-footprint proxy.
Retrieved 2026-06-25Social profile.high
X - Sean Grove
x.com
Founder public social footprint.
Retrieved 2026-06-25Social profile.medium
GitHub - Sean Grove
github.com
Founder public technical footprint.
Retrieved 2026-06-25Public code profile.high
GitHub - Linzumi
github.com
Public product-organization repository signal.
Retrieved 2026-06-25Public code organization.high
Y Combinator profile - OneGraph
ycombinator.com
Prior company and acquired status.
Retrieved 2026-06-25Accelerator company page.high
Y Combinator profile - Zenbox
ycombinator.com
Prior YC company and inactive status.
Retrieved 2026-06-25Accelerator company page.high
Netlify acquires OneGraph
netlify.com
OneGraph acquisition and founder track record.
Retrieved 2026-06-25Company acquisition announcement.high
Digg launch coverage
digg.com
YC ecosystem launch amplification.
Retrieved 2026-06-25Press and social aggregation page.medium
Gartner AI code-assistant adoption forecast
gartner.com
AI coding-assistant adoption forecast.
Retrieved 2026-06-25Analyst press release (Gartner).high
Stack Overflow Developer Survey 2025 - AI
survey.stackoverflow.co
Developer AI-tool and agent-adoption data.
Retrieved 2026-06-25Developer survey (Stack Overflow).high
BLS software developers outlook
bls.gov
U.S. software-developer employment denominator.
Retrieved 2026-06-25Government statistics (BLS).high
Microsoft 2024 annual report
microsoft.com
GitHub Copilot paid-subscriber proxy.
Retrieved 2026-06-25Public-company filing (Microsoft).high
Microsoft 2025 annual report
microsoft.com
GitHub Copilot user proxy.
Retrieved 2026-06-25Public-company filing (Microsoft).high
GitLab FY2025 results
ir.gitlab.com
Developer-platform revenue and margin benchmark.
Retrieved 2026-06-25Public-company release (GitLab).high
Atlassian FY2025 annual report
s206.q4cdn.com
Collaboration-platform revenue, customer count, and AI-agent strategy.
Retrieved 2026-06-25Public-company filing (Atlassian).high
IDC AI coding assistants report
services.google.com
Enterprise AI coding-adoption and productivity context.
Retrieved 2026-06-25Analyst report with sponsor caveat (IDC).high
AgentsRoom
agentsroom.dev
Adjacent agent-orchestration competitor.
Retrieved 2026-06-25Competitor company page.high
OpenACP
openacp.ai
Open-source chat bridge and local-agent substitute.
Retrieved 2026-06-25Competitor company page.high
Crewly
crewlyai.com
Adjacent orchestration user interface.
Retrieved 2026-06-25Competitor company page.high
Codezal
codezal.com
Adjacent orchestrated coding workspace.
Retrieved 2026-06-25Competitor company page.medium
Devin Desktop / Windsurf
windsurf.com
Desktop fleet-management alternative.
Retrieved 2026-06-25Competitor company page.high
GitHub Copilot
github.com
Incumbent AI coding assistant and pricing benchmark.
Retrieved 2026-06-25Competitor company page.high
Claude Code
code.claude.com
Incumbent coding agent and Slack/team entry-point signal.
Retrieved 2026-06-25Competitor company page.high
OpenAI Codex
openai.com
Incumbent multi-agent coding workflow.
Retrieved 2026-06-25Competitor company page.high
Cursor
cursor.com
IDE-agent incumbent and multi-agent collaboration pressure.
Retrieved 2026-06-25Competitor company page.high
Devin
devin.ai
Hosted agent substitute.
Retrieved 2026-06-25Competitor company page.high
Cline
cline.bot
Local coding-agent substitute.
Retrieved 2026-06-25Competitor company page.medium
Aider
aider.chat
CLI coding-agent substitute.
Retrieved 2026-06-25Competitor company page.medium
OpenHands
github.com
Open-source coding-agent platform substitute.
Retrieved 2026-06-25Public code repository.medium
vibe-kanban
github.com
Open-source agent-orchestration substitute.
Retrieved 2026-06-25Public code repository.medium
Claude Squad
github.com
Open-source multi-agent terminal substitute.
Retrieved 2026-06-25Public code repository.medium

Access limitations: This research did not have access to Linzumi's private product analytics, billing data, cap table, customer references, source code, security architecture, SOC 2 materials, or employment records. Public search checks also did not surface customer case studies, public job postings, Product Hunt, or Hacker News product-launch records. Those limitations mean the report can assess founder signal, category timing, public positioning, and public risk surfaces, but cannot verify retained usage, revenue quality, gross margin, full team coverage, or enterprise trust readiness.

Public disclaimer: This is an independent public teardown based solely on publicly available information retrieved on 2026-06-25. It is research support, not investment, legal, tax, or financial advice. Company-provided statements are labeled as claims, not verified facts. Linzumi did not participate in or review this report. Errors and omissions are possible; final decisions remain with the reader.

What The Company Does#

Key Takeaways#

What Makes This Potentially Fundable#

The Four Holes To Close Before Fundraising#

Decision Snapshot#

IC Disagreement Map#

Scenario Range#

What Investors Will Test#

Claim Reconciliation: Inconsistencies Investors Will Catch#

Diligence Findings: Issues To Fix Before You Raise#

Data Room Readiness Checklist#

First-Call Agenda For The Startup#

Business Model And Revenue Signals#

Revenue Quality Checklist#

Funding And Ownership Context#

Founder, Team, And Related Entities#

Founder And Team#

Founder Competency Coverage#

Public Professional Footprint#

Founder-Market Fit Read#

Related Entities And Founder-Associated Companies#

Traction#

Customer Status Table#

Traction Signals#

Hiring And Org Momentum#

Traction Quality Read#

Competitive Landscape#

Category Visibility Snapshot#

Pricing And Competitive Benchmark#

Competitive Wedge#

Risks And Open Questions#

Pre-Mortem: The Most Likely Obituary#

Decision-Critical Unknowns#

Diligence Questions#

First Call#

Follow-Up#

Kill Criteria#

Double-Down Criteria#

Founder Action Plan#

Decision#

How We Would Miss This One#

Source Log#