Join us

Build with us

Medusa gives lean investment teams the intelligence of top-tier hedge funds, on the stack they already run. We are looking for a technical cofounder to lead engineering as the company grows, and for senior AI engineers who want to ship in front of real clients.

What we are building

Medusa is a three-layer platform. It ingests the data a firm already has, in Excel, cloud storage, broker files and market data feeds, reconciles it into one view of the portfolio in the team's own definitions, and runs AI agents on top. The first agent, investment risk, is live. Watch the demo, and the short film of where it goes: an agent at the investment committee table.

A general model on its own cannot answer a portfolio manager in the firm's own shorthand. We build the harness that lets it: the context, tools, skills and evaluation around the model, with every figure cited to its source. We call that discipline context engineering. The platform is model-agnostic by design.

Medusa was founded in January 2026 by Petr Merkuryev after 22 years at the intersection of finance, technology and data science, including 11 inside hedge funds. The workflows we automate are the ones he ran himself.

Where we stand

A working platform with a live risk agent: factor exposures, stress tests, limit checks and daily briefings.

Revenue from launch through hands-on client deployments.

The team today: the founder, a fractional CTO who owns architecture and security, contract AI engineers as the work requires, and two industry advisors, a hedge fund founder and CEO and a PhD quant researcher.

Open roles

Technical cofounder

One seat, with a paid way in. Start part-time, one to two days a week, paid, building alongside the founder and our fractional CTO. Grow into the technical cofounder role, with founder-level equity, as the fit proves itself on real work. Structure and vesting depend on profile, timing and commitment.

You would own

Platform engineering from MVP to production: the agent harness, the reconciled data layer, evaluation.
Security and deployment for hedge fund clients, together with our fractional CTO, who owns the security posture today. Private cloud and on-premises options matter in this market.
Building the engineering team as the platform work grows.

Who fits

Senior engineering leadership with production systems shipped and operated, not advisory-only.
Hands-on with LLM agent systems: tool use, structured outputs, evaluation. Or the ability to get deep fast.
At ease in data-heavy, security-sensitive environments. Capital markets exposure is a plus, not a requirement.
Builder temperament: small team, high autonomy, AI-assisted development, real clients within weeks.

Senior AI engineers, contract

Scoped deliverables on the platform and on client deployments, contracted through Deel, remote, European hours preferred. Typical work: a new agent tool with its evaluation, a data-source integration, packaging a deployment for a client's environment.

Who fits

Python and LLM agent engineering in production. Our stack: Python, DuckDB and pandas, Next.js and React, Azure and private-cloud deployments.
Comfortable working from a spec you helped write, and proving the result with an evaluation before it is shown to a client.
Available for a first scoped piece within weeks, with the option to continue.

How we work

Three rules shape the product and the way we build it: agents see the portfolio the way the team sees it, they help the team decide and never decide for it, and everything they say is verifiable, every figure cited to its source. Small team. The founder is in every engagement, and evaluation comes before the demo.

How to apply

Write to Petr at petr.merkuryev@medusaip.ai with the subject "Join us: cofounder" or "Join us: engineer". Send a few lines on what you have shipped and operated, and a link to something you built. No cover letter needed. Every message gets a reply.

Zurich, remote-friendly.