The opportunity

A venture-backed engineering team in New York is hiring the person who will own technology through its next chapter. The current technical leader built the foundation that got the company here and is deliberately stepping toward a lighter, later-stage role. This is his successor: the technical face of the business with customers, the architect who turns one-off enterprise deployments into something repeatable, and the leader who lifts the engineering bar without burying the team in process.

What the company does

The hard part of enterprise data is the material no software can read — contracts, financial paperwork, scanned files that arrive as noise. The company turns that noise into clean, trusted data, then builds the software and agents that run on top of it. Its first market is the portfolio companies of institutional investors, a segment where one buyer can mandate adoption across dozens of businesses that all share the same problem.

Three things make the work distinctive:

• Accuracy is the product. Every value traces back to the document it came from and stands up to audit. That provenance is what lets a customer put the output into production rather than into a demo.
• Engineers go to the customer. The team embeds its own engineers on site, scopes the real problem inside a week, and stays accountable through to production.
• The work compounds. Each engagement gets productized into reusable pieces, so every deployment makes the next one cheaper. A general-purpose model cannot replicate that, because the edge comes from real customer problems and the review that follows them.

Traction

Backed by a top-tier accelerator and growing quickly — from zero to meaningful recurring revenue inside the first year, on the back of a lean team of under fifteen in Manhattan. A majority of the largest institutional investment funds in the world are already customers, and their deal teams make warm introductions directly to the operators running the portfolio. The founding story is not academic: the CEO lived this exact document-and-reconciliation pain from the buyer's side before building the fix.

What you will own

• Technical direction. Set the architecture for the agentic platform and the data layer beneath it: high-accuracy extraction, evaluation harnesses, schema mapping, agent orchestration, and integration into customer systems of record.
• Customer-facing design. Sit with sophisticated institutional customers, separate what they asked for from what they actually need, and convert that into a technical plan.
• Hands-on work. Roughly a third to half your time is real technical work at the start. That share drops over time as architecture, review, and leadership take over.
• The delivery function. Build and lead the embedded-engineering practice that carries customer engagements end to end. This is what customers are paying for.
• Hiring. Grow the engineering team from its current handful. Whether that lands at six or forty depends on customer pull and financing.
• Product partnership. Work with the product co-founder on roadmap, sequencing, and build-versus-buy.
• Founder-level scope. A seat in the conversations that matter — sales, customer success, technical diligence, hiring, and fundraising.

The first three months

Month one. Earn the trust of a strong engineering team, audit the existing architecture, and lead the design of a multi-source ingestion agent for a major customer rollout: pulling from shared drives and email, mapping schemas, preserving provenance, and handing clean data off downstream.

Month two. Turn that single deployment into a reusable pattern — connectors, extraction and evaluation harnesses, versioning rules, citation UX, and playbooks the delivery team can run again elsewhere.

Month three. Own the room with customer CTOs, security leads, finance chiefs, and operators. Make the build-versus-configure calls. Decide which two to four engineers to hire next.

By the end of the quarter the company should feel it has a technical owner at cofounder level: customers trust you, engineers learn from you, and the founders are no longer the bottleneck on technical judgment.

Who this fits

• Cofounder-shaped. Possibly an ex-founder, a former CTO, a founding engineer, or a staff or principal engineer who wants ownership at cofounder level.
• Both pedigrees. The pattern that works is several years inside a genuinely elite engineering organization, plus real startup or founding-engineer mileage.
• Still technical today. You can code, debug, review architecture, and use modern AI coding tools with judgment. The team has to be able to learn from you.
• Comfortable in front of customers. You can hold your own with demanding enterprise buyers and turn ambiguity into execution.
• Scarred by enterprise data. You know exactly why document AI dies in production: OCR noise, drifting schemas, tables, citations, evaluations, permissions, and the handoff into a system of record.
• A leader, not an administrator. You can manage strong engineers without importing heavyweight process.
• Present. This is in-person in Manhattan, at a founder's operating tempo rather than a settled corporate rhythm.

What this is not

• Not a pure research or ML individual-contributor seat — you lead people and own outcomes.
• Not a late-stage VP of Engineering post — there is no large organization to administer.
• Not remote, and not advisory. You need to be in the room.
• Not generic enterprise data work. The edge is document-heavy mess, provenance, evaluation, and shipping into live customer systems.

Helpful extras

• Embedded or forward-deployed engineering experience, or solutions architecture at a company known for that model.
• Exposure to private markets, fintech, enterprise AI, document AI, OCR, or workflow automation.
• A track record of people following you between companies.
• Visible or open-source work in AI agent tooling.
• Visa transfers can be considered for exceptional candidates already in New York, or seriously committed to moving.

Compensation

• Base. Around $200,000, with room to flex for an exceptional candidate.
• Equity. Up to 10% — a genuine founder-level grant.
• Benefits. Full health, dental, vision, and 401(k).
• Kit. Best-in-class hardware and tooling.

How the process runs

• Three or four conversations with the founders and team.
• A two-day working session on site. This is not a canned coding test — it is mutual diligence on how you think, build, lead, and handle the company's actual problems.
• A meal or coffee with the team before any decision.
• For the right person, first call to decision can happen in about a week.