Founding AI Platform Engineer

Leorna is building the big-data platform that gives AI agents persistent long-term memory — a "forever memory" layer so agents can store, retrieve, and reason over everything they've learned instead of starting from scratch every session.

We're very early, pre-funding, and moving fast. There's no platform team to hand things off to, no legacy system to work around, and no one above you owning the architecture. You'd be building it.

THE ROLE

We're looking for a Founding AI Platform Engineer to own the technical core of the platform and the agentic tooling that builds it. This is a hands-on founding role with unusual breadth. You'll make the architecture calls and then write the code. You'll pick the stack, stand up the infrastructure, build the automation, and set the standards.

If you've been the person at a startup who ends up owning whatever's broken and most important that week, this is that job with a real title attached.

WHAT YOU'LL DO

Architect and build the memory platform

• Design and implement the memory storage and retrieval layer: vector search, embeddings, context management, and the data pipelines that feed them
• Choose the models, infrastructure, and data stores the product runs on, and document the tradeoffs
• Stand up evaluation, benchmarking, and reliability practices so agent memory stays accurate and fast under load
• Take features from prototype to hardened production, often in the same week

Run our agentic development stack

• Set up, configure, and operate OpenClaw, Hermes, Claude Code, and OpenAI Codex as first-class parts of our development workflow
• Build and maintain MCP servers, skills, agent configs, and local agent gateways the rest of the team uses daily
• Debug the unglamorous parts: config files, connection failures, tool permissions, context management, and make the setup reproducible for new hires

Automate everything worth automating

• Design and build automation for CI/CD, testing, deploys, data pipelines, code review, and internal ops
• Replace recurring manual work across engineering and business functions with reliable agentic or scripted workflows
• Keep the automation observable and maintainable, not a pile of brittle scripts

Own the repos and engineering standards

• Manage multiple repositories, including structure, branching strategy, and cross-repo dependencies
• Handle real Git maintenance: rebases, history cleanup, submodules or monorepo tooling, tagging and release management, hooks, and access control
• Set code review standards, CI gates, security practices, and documentation norms
• Help shape technical hiring and the roadmap as we grow

WHAT WE'RE LOOKING FOR

Must-haves

• Deep experience architecting and building production systems. You've owned systems end to end, not just designed them
• Real working knowledge of OpenClaw, Hermes, Claude Code, and OpenAI Codex
• Strong AI/ML engineering background: LLMs, embeddings, vector search, retrieval, context management, and agentic systems
• Serious automation and workflow engineering skills: CI/CD, orchestration, scripting, infrastructure as code
• Advanced Git and multi-repo experience, including the maintenance work most engineers avoid
• Comfort operating with high ambiguity, high ownership, and no playbook
• You currently hold and actively use a premium AI subscription (Claude Max, ChatGPT Pro, or equivalent at the $200/month tier)
• Located in the United States

Nice to have

• Early technical hire at a pre-seed or seed-stage startup
• Distributed systems, large-scale data infrastructure, or database internals
• Open-source contributions to agent tooling, MCP servers, or developer infrastructure
• Security and IT administration experience

TERMS: PLEASE READ BEFORE APPLYING

• This is a founding-team, equity-only role right now. There is no cash compensation until we close funding. After funding, this converts to salary plus continued equity.
• Fully remote. We're distributed and async-friendly.
• High dedication, fast pace. We're pre-funding and building against a real window.
• You must hold an active Claude or OpenAI Codex subscription and use AI tooling as a core part of how you work.

If equity-only doesn't work for your situation right now, we completely understand, and we'd still like to talk when we've raised.

HOW TO APPLY

Apply here, or grab a time directly on my calendar: https://cal.com/mnylif/interview

Tell me about a system you architected and built yourself, and how you're using agentic tooling in your own workflow today.

• Michael, Founder