Key Skills

Job Description

Python AI Engineer (Junior) | DODO OÜ About Us Data Oriented Defence Operations (DODO OÜ) is a cutting edge Cybersecurity boutique based in Tallinn, Estonia. We operate at the forefront of Deep Tech, and yes, it’s time to say it openly: we proudly emerged from the NATO ecosystem. We build advanced Defence and Security solutions and push the boundaries of autonomous offensive security. The Role We are looking for a Python AI Engineer (Junior) to join our development team and work on ARES, our autonomous red teaming platform powered by Agentic AI. This is a part time position of about 22 hours per week, fully remote, with flexible scheduling. It’s designed to fit alongside a Master’s program or other commitments. Given where our product is, we want someone who has already spent real time building agent systems, ideally with LangChain and LangGraph, and who is ready to contribute to a live, complex agentic infrastructure early on. You’ll be closely mentored by senior profiles in AI and Cyber Operations, so the learning curve is steep and well supported. We just want it to start from solid hands on ground. Eligibility Because of the nature of our work in Defence and Security, we can currently only consider candidates who hold European (EU/EEA) citizenship and live in Europe. We appreciate the interest from everyone else, but unfortunately we can’t move forward with those applications at this stage. Compensation & Company Stage: Please Read Carefully We want to be transparent about where DODO stands today, so you can make an informed decision. We have not closed our first investment round yet, and we are working to close it. Concretely: The role is a 6 month engagement at about 22h per week. The first 3 months are a trial period. Compensation is €3,800 to €4,200 in total, depending on experience, paid in two tranches. If the funding round closes, the role can convert to a full time paid position with a stock option plan, on terms set out in the Memorandum of Understanding we sign at the outset. If the round does not close, the engagement ends after the 6 months with the agreed compensation paid, and the move to a full time role and stock options would not go ahead. We understand the compensation is modest and the structure won’t suit everyone, and that’s entirely reasonable. If it’s not the right moment, we’d welcome hearing from you again once we’re funded. What You Will Do Agentic Backend Development: Design and optimize agent architectures using LangChain and LangGraph, supporting our CTO and the engineering team. Agentic RAG Pipelines: Build retrieval augmented generation pipelines that let agents query knowledge bases, vulnerability databases, and prior engagement data to inform autonomous decisions. Observability & Evaluation: Instrument agent pipelines with Langfuse for tracing, evaluation, and debugging of autonomous agent behavior. Structured Data & Validation: Use Pydantic to define robust schemas for agent inputs and outputs, tool calls, and communication between agents. Memory & State Systems: Build and maintain memory layers (vector stores, relational and database backed memory) so agents keep context across sessions and tasks. Target & Vulnerable Environment Creation: Build automated scenarios and vulnerable targets that our offensive AI attacks during QA and UAT phases. Cross Functional Bridging: Act as the technical bridge between the Agentic AI development team and the operational Cybersecurity team for test integration. What We Are Looking For Hands on experience building agentic AI systems: projects you’ve actually built and can talk us through, including the architecture you chose, what worked, and what you’d do differently today Practical experience with LangChain and LangGraph: structuring graphs, managing state across nodes, handling tool calling, and debugging agents that loop or stall Excellent Python Solid understanding of RAG: embeddings, chunking strategies, vector similarity search, and retrieval grounded generation, ideally with experience tuning a pipeline yourself Pydantic (or similar) for data validation and schema design Comfort with databases and state persistence patterns, both vector stores and relational or document stores Master’s degree, enrolled or completed, in Computer Science, Cybersecurity, AI, or a related field Ability to work autonomously on goal oriented tasks in a fully remote setup, managing your own 22h per week Genuine interest in offensive security, cloud architectures, and applied AI Nice to Have Python backend experience: building and consuming APIs (FastAPI, Flask, or similar), async programming, and services that hold up in production Langfuse or similar LLM observability and evaluation tools Exposure to offensive security tooling, CTFs, or pentesting fundamentals Experience with multi agent orchestration patterns beyond LangGraph Familiarity with Docker and cloud deployment on any major provider What We Offer Paid 6 month engagement: €3,800 to €4,200 in total depending on experience, paid in two tranches. The first 3 months are a trial period. 100% fully remote, flexible part time (within Europe): about 22h per week, organized around your own schedule Expert Mentorship: Daily collaboration with top tier industry experts in AI and Cyber Operations Deep Tech Impact: Direct involvement in a Deep Tech project in Defence and Security, with real impact on our technology roadmap Career Growth: A path to a full time role with a stock option plan if the funding round closes, and growth into key architectural roles as we scale How to Apply Apply with your CV. If your profile includes agentic AI projects (personal, academic, or professional), make sure they’re visible on your CV or LinkedIn, since that’s what we’ll look at most closely.

Core Responsibilities

Design and optimize agent architectures and RAG pipelines, instrument them for observability and evaluation, and build robust schemas and memory systems. Create vulnerable test environments and coordinate integration between the Agentic AI development and cybersecurity teams.

Requirements

Candidates should have hands-on experience building agentic AI systems, practical LangChain and LangGraph skills, strong Python, and a solid understanding of RAG, data validation, and database persistence. A master's degree in Computer Science, Cybersecurity, AI, or a related field is sought, along with autonomous remote-work capability and genuine interest in offensive security, cloud architectures, and applied AI.

Benefits

• Flexible Schedule
• Remote Work
• Expert Mentorship
• Potential Full-Time Role
• Potential Stock Options
• Career Growth