Is Jarmin profitable in 2026?

No revenue, ARR, or profitability data has been disclosed for Jarmin. Its own launch materials (Nov 10, 2025) describe only illustrative use cases with no named customers, and a PitchBook profile from 2026 lists investors but no confirmed funding or financial figures.

Could AI replace Jarmin?

Jarmin's own business is building AI to replace human roles like ML engineers and paralegals, but the same frontier labs (OpenAI, Anthropic, Google) it depends on, plus incumbents like Salesforce Agentforce, ship comparable agentic features natively, leaving Jarmin's thin orchestration layer exposed to being replicated or bypassed.

Jarmin is privately owned by its founders and is a member of Y Combinator's Fall 2025 (F25) batch, based in San Francisco. No acquisition, parent company, or ownership change has been reported; a PitchBook profile lists three unnamed investors as of 2026.

Has Jarmin raised outside funding?

No confirmed seed or Series A round is publicly disclosed for Jarmin beyond its standard Y Combinator backing as an F25 batch company. A PitchBook profile from 2026 references three investors, but names, amounts, and valuation are not verifiable in public sources.

What does Jarmin actually sell?

Jarmin sells autonomous AI agents it calls AI employees for roles including ML Engineer, Financial Analyst, Account Executive, Compliance Analyst, Data Analyst, Project Manager, and Paralegal. Its flagship ML Engineer product launched via Launch YC on Nov 10, 2025, with pricing undisclosed.

YC Fall 2025 · United States

Jarmin

Startup selling AI agents that do human jobs like analyst, paralegal, sales rep.

Little protection left

Assessed 27 July 2026 · Evidence as of 27 July 2026

Not much stands between this business and its pressures.

On the recordSince 27 July 2026 · Checkable from H2 2026 · Every call on the record Checkpoint: Outcome of Jarmin's first Founding Sales/Account Executive hire and any funding round revealing horizontal vs vertical focus

The call

Jarmin sells AI workers into a category where a peer lost most of its customers in months and the biggest software firms already ship the same idea natively.

Jarmin sits between customers and AI models like OpenAI or Claude, trying to coordinate them like a traffic controller. The problem is that Salesforce, Microsoft, and Google are all building the same capability directly into their own systems where customers already work, and those companies have way more customers and control. The one thing that could matter: if Jarmin can show it actually helps companies finish projects that fail elsewhere, and if it builds a real sales team instead of relying on the founder, then it might survive the next two years.

Why

AI pressure: OpenAI, Anthropic, Google frontier models. The same frontier models Jarmin wraps could ship agent orchestration directly, cutting out the wrapper layer. Could AI do the core job? Yes, likely.

Bigger players: Salesforce Agentforce, Microsoft 365 Copilot agents, Google Gemini for Workspace, ServiceNow, Moveworks. These platforms already sell agentic roles baked into the systems companies already run their business on. Could a giant copy it? Yes, easily.

What protects it. None apparent: no named customers, no disclosed retention data, no proprietary integration or regulatory footing.

What would change this call

Jarmin announces a seed or Series A round and states its target market. A horizontal 'AI workforce' pitch signals the same doorway bet that failed for 11x; a single regulated vertical signals a narrower, more defensible strategy.
Jarmin's Founding Sales/Account Executive role (posted Jul 2026) is filled and produces named customer wins within 24 weeks. Shows whether Jarmin can move from founder-led sales to structured GTM with real traction.
Salesforce, Microsoft, or ServiceNow announce new agentic role-based features (e.g., compliance or paralegal agents) matching Jarmin's role list. Would show incumbents closing the exact gap Jarmin is trying to occupy.

Key numbers

Team size: 5 people (Jul 2026)
Product roles expanded from 1 to 7 in about 8 months (mid-2026)
Best tested AI agent completed only 2.5% of projects (Mar 1, 2026)

Named risks

No proprietary data, regulatory cover, or exclusive integrations disclosed to defend against replication. Severity: high
Major platforms (Salesforce, Microsoft, Google, Moveworks) ship comparable agentic features with far greater distribution. Severity: high
Category shows early disillusionment: peer 11x lost 70-80% of customers and best AI system completed only 2.5% of projects in 2026. Severity: medium
Extremely small team (1-5 to 5 people) with no disclosed funding beyond standard YC backing as of Jul 2026. Severity: medium
No named customers or independent traction metrics; only illustrative marketing use cases as of Nov 2025. Severity: medium

Rivals, as read in July 2026

Rival	Direction	Evidence
Salesforce Agentforce	losing to	Ships autonomous task-execution natively inside Salesforce systems as of 2026
Microsoft 365 Copilot agents	losing to	Ships autonomous task-execution inside Microsoft enterprise systems as of 2026
Google Gemini for Workspace	losing to	Ships autonomous agents inside Google enterprise systems as of 2026
Moveworks	losing to	Reasoning Engine ships multi-step agentic HR/IT workflow execution, Jun 10, 2026
11x	contesting	Peer AI-employee vendor reportedly lost 70-80% of customers in 2026, showing category-wide churn risk

Questions people ask

Is Jarmin profitable in 2026?: No revenue, ARR, or profitability data has been disclosed for Jarmin. Its own launch materials (Nov 10, 2025) describe only illustrative use cases with no named customers, and a PitchBook profile from 2026 lists investors but no confirmed funding or financial figures.
Could AI replace Jarmin?: Jarmin's own business is building AI to replace human roles like ML engineers and paralegals, but the same frontier labs (OpenAI, Anthropic, Google) it depends on, plus incumbents like Salesforce Agentforce, ship comparable agentic features natively, leaving Jarmin's thin orchestration layer exposed to being replicated or bypassed.
Who is Jarmin's biggest competitor?: Jarmin competes against enterprise incumbents shipping agentic AI directly into systems they already own, including Salesforce Agentforce, Microsoft 365 Copilot agents, Google Gemini for Workspace, and Moveworks, all noted as actively shipping comparable autonomous task-execution capability through 2026.
Who owns Jarmin?: Jarmin is privately owned by its founders and is a member of Y Combinator's Fall 2025 (F25) batch, based in San Francisco. No acquisition, parent company, or ownership change has been reported; a PitchBook profile lists three unnamed investors as of 2026.
Has Jarmin raised outside funding?: No confirmed seed or Series A round is publicly disclosed for Jarmin beyond its standard Y Combinator backing as an F25 batch company. A PitchBook profile from 2026 references three investors, but names, amounts, and valuation are not verifiable in public sources.
What does Jarmin actually sell?: Jarmin sells autonomous AI agents it calls AI employees for roles including ML Engineer, Financial Analyst, Account Executive, Compliance Analyst, Data Analyst, Project Manager, and Paralegal. Its flagship ML Engineer product launched via Launch YC on Nov 10, 2025, with pricing undisclosed.

Timeline

2026-07-22Jarmin posts its first Founding Sales/Account Executive job listing
2026-07-21Jarmin lists AI Engineer, Fullstack Engineer, and Backend Engineer job openings
2026-06-30HBR piece frames 2026 as pivotal year for agentic AI startups versus incumbents
2026-03-09Fondo covers Jarmin.ai launch of 24/7 Machine Learning Engineer employees
2026-02-27Industry report says peer 11x lost 70 to 80 percent of customers in 2026
2025-11-10Jarmin launches autonomous ML Engineer AI employee via Launch YC
2025Jarmin founded and joins Y Combinator Fall 2025 (F25) batch in San Francisco

Revision history

27 July 2026

First assessed. Little protection left. The text above is fixed as of this date and is not edited afterwards.

In the same band

Exonic Openroll Semiotic