AI Platform Engineering Lead

Obin AI • New York, NY

3d

About The Position

This is the founding platform engineering leadership role. You will own the paved road: the scaffolding, tooling, isolation model, and release machinery that lets engineers ship compliant agent workflows in a day instead of weeks — and, before long, let a customer's own engineers do it without us in the room. You will lead a small, high caliber and capable team. A meaningful part of this job is establishing the engineering standards.

Requirements

• Built and operated a multi-tenant platform that external developers built on — SDK or API surface, tenant isolation, gated deploys, the whole path.
• Shipped in a regulated environment and can hold your own on SOC 2, audit trails, access control, and model risk without treating them as someone else's problem.
• Converted services delivery into product before. You know which bespoke work becomes a platform primitive and which is a one-off you should refuse to generalize.
• Led engineers directly, including engineers earlier in their careers, and you set standards by writing things down rather than by being in every conversation.
• Opinionated about developer experience and can defend those opinions with what happened last time.

Nice To Haves

• Internal developer platform, paved road, or platform-as-a-service ownership at a company that sold to enterprises.
• Said no to a customer request in order to protect a platform abstraction — and can explain how you made that call.
• Introduced written decision-making to a team that ran on hallway conversation, and it survived contact.
• Comfort with AI agent systems is welcome but not required.

Responsibilities

• Own the golden path: scaffolding, templates, and defaults so a new solution stands up with backend, ingest, RBAC, and audit packaging already handled — and the engineers write only agent logic.
• Own the shared tool catalog: Make tools first-class, versioned, referenced objects rather than copy-paste.
• Own versioning and promotion lanes: Agents and playbooks move local → shared → production through auditable, gated CI/CD.
• Own the compliance harness: Provable auditability and governance for every agent run: who changed what, which version executed, which sources grounded which claim.
• Own testing infrastructure: A skill and tool testing suite, regression gates, and shadow A/B in production environments.
• Own multi-tenancy done properly: Logical isolation across the control, data, execution, and playbook planes — strong enough that a customer's security review passes without us resorting to dedicated per-tenant infrastructure.
• Own the customer-facing developer experience: SDK, API gateway, sandboxed evaluation, trace-level debugging, and egress webhooks that a bank's own quant engineers can safely build against.
• Own a repeatable vendoring cycle: Make shipping sanitized codebases into customer environments a supported, IP-defensible process.
• Establish engineering standards that stick: Written specs, clear approval rights, real separation of concerns, and a team that knows the difference between deliberate technical debt and accidental technical debt.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior
• Education Level: No Education Listed