PhD-level Engineer

Introduction

Calypso Ventures GmbH is an AI company based in Berlin, Regensburg, and London, founded by two Oxford graduates. We build AI optimization software for global commodity trading, specifically for LNG, crude oil, and power. Our clients include some of the world's largest energy and trading companies. Our technology not only enables smarter trading decisions but also reduces CO2 emissions equivalent to what 330 million trees absorb annually. We are expanding to Singapore and growing fast.

The Pitch

Neo is our AI trading analyst. A fully autonomous agent that our clients use every day across our products X-LNG, X-Crude, and More. Today, Neo runs entire optimizations from natural language, analyzes results, compares scenarios, and generates executive-ready outputs directly from real trading data. PowerPoints, PDFs, Excel reports, podcasts, videos.

We want Neo on every single trading desk on earth. We are looking for a PhD-level engineer to take full ownership of building the system around it. This is not about training new foundation models. We use the best models that already exist. Your work is about everything that turns a model into a reliable production agent: the architecture, the tools, the memory, the orchestration, the APIs, the evaluation, the integration across six products. It is real software engineering on top of state-of-the-art AI.

Tasks

• Build and extend the agent system that powers Neo across all our products
• Design tool integrations, memory systems, and orchestration logic that let Neo act autonomously
• Develop the APIs, backend services, and infrastructure that make agents reliable at scale
• Evaluate and integrate new LLMs, SDKs, and agent frameworks as the landscape evolves
• Design and run evaluation pipelines to measure agent reliability on real trading workflows
• Work directly with traders to translate their workflows into agent capabilities
• Collaborate closely with the founders and the product team on the long-term AI roadmap
• Write production Python, ship features, review code, and own quality end to end

Profile

• A PhD (or strong Master's with deep industry experience) in Computer Science, Machine Learning, Engineering, or a related field
• Strong software engineering skills in Python. You write production code and own it end to end
• Solid understanding of modern LLMs and the current agentic landscape (tool use, memory, evaluation, orchestration). You do not need to train foundation models yourself, but you know how they behave
• Hands-on experience with agent frameworks or the ability to build them from first principles
• Comfort with real software engineering work: backend services, APIs, databases, deployment, monitoring. This is not a pure research role
• The ability to think deeply and solve hard problems. You break down complex systems, build structured solutions, and ship
• Strong English skills; German is a plus

We offer

• Full ownership of a product that real traders already rely on. Neo is live today, and your work ships to production every sprint
• Direct access to the founders and a team of around thirty mathematicians, physicists, engineers, and computer scientists
• Possibility to travel to clients in London, Singapore, Houston, Norway, Geneva, the Middle East, and Australia
• Relocation support to make your move to Berlin, Regensburg, or London easier
• Flat hierarchies, direct access to founders, and a team that actually likes working together
• Free drinks available at any time
• A wide range of professional development opportunities
• Sports activities for you to enjoy at your leisure
• Comprehensive onboarding to help you settle in quickly
• Your success is rewarded with profit-sharing
• Your ideas are welcome and valued