Build the agents thatproduction can actually trust.

We are a small team building a hyper-converged AI agent platform for prod — Kubernetes RCA, autonomous remediation, and an audit ledger you can hand to your auditor. We are hiring opportunistically for early roles and welcome design partners. If anything below sounds like you, send a note.

Remote-first · US Pacific or US Eastern overlap preferred · no relocation required.

What we look for

Hiring filters that matter to us more than years of experience or the right keywords on a CV.

  • Code-grounded marketing

    Every claim on the landing page traces back to code. We do not invent benchmarks.

  • Evidence over confidence

    Every AI answer cites the log lines, metrics, or commits it used. The agents do not get to bluff. Neither do we.

  • Approval-gated execution

    An autonomous agent that ignores humans is a liability. Ours always asks before touching prod. We hire people who think the same way about their teammates.

  • Small team, high trust

    Few people, shared context, no PR-of-the-week culture. You will own surfaces end to end and answer for them.

Where we are hiring (loosely)

No formal job postings yet — we read every email and reply. If your background does not fit any of these but you are excited about the problem space, write anyway.

  • Engineering

    Express Interest

    Founding Engineer — Backend / Agents

    Own the master orchestrator and the specialist agents that fan out to Kubernetes, vector stores, and code repos. Strong Python, comfortable with distributed systems, allergic to fragile abstractions.

  • Engineering

    Express Interest

    Founding Engineer — Full-Stack

    Bring the operator console, the conversation surface, and the audit ledger to life. React + TypeScript on the front, FastAPI on the back. Care about UX. Read tokens.css before you commit.

  • Engineering

    Express Interest

    Site Reliability Engineer

    Your stack is the blast-radius simulator and the SLO guardian. You have lived through bad rollouts and have opinions about why incident-response tools never quite work in practice. Help us make this one work.

  • Founding Designer

    Define how an operator-facing AI product should feel — when it is right, when it is uncertain, when it has just made a change you need to reverse. Tokens, motion, density, trust.

  • Design Partner / Customer Zero

    Not a job posting — an open call. If you run a Kubernetes-heavy stack and want to put an autonomous RCA + remediation agent in front of your prod with full audit, we want to talk. Free for the first three months in exchange for honest feedback.

Don't see your role?

We are early. Email hi@p95.ai with a paragraph about what you would build with us and a link to anything you have shipped that you are proud of. Cover letters not required.