Iryss AI is a leadership intelligence platform that helps private equity firms predict which executives will deliver. We transform management due diligence from slow, subjective gut-feel into fast, data-driven prediction — analysing signals from assessments, interviews, and open-source data to benchmark leadership quality. Each interaction compounds our intelligence, allowing the system to learn continuously from new data and outcomes.
We’ve already built and demoed a working prototype with several private equity firms who’ve expressed strong interest in paid pilot programs once the full product is ready. You’ll join at a pivotal point — turning early traction into a scalable platform that defines a new category in leadership intelligence.
Role: Founding Generative AI Engineer (LLM Systems & Reinforcement Learning)
This is a full-time remote role for a Founding Generative AI Engineer.
You’ll be responsible for designing, building, and deploying the AI systems that power Iryss — from LLM-driven insights and benchmarking pipelines to feedback and evaluation systems that make the platform smarter with every use.
You’ll work directly with the founder to bring the Iryss vision to life: a continuously learning leadership intelligence engine trusted by investors and CEOs.
Key Responsibilities
* Architect and implement LLM-powered workflows using OpenAI, Anthropic, and other APIs (e.g. summarisation, analysis, benchmarking).
* Develop retrieval-augmented generation (RAG) pipelines connecting structured/unstructured data (Supabase, Pinecone, Chroma).
* Design and maintain evaluation systems — to assess LLM output quality, bias, and reliability.
* Build a reinforcement learning feedback loop, where user interactions and outcome data improve future model performance.
* Collaborate with the founder on data ingestion, prompt design, and multi-agent orchestration (e.g. strategy, analysis, reporting agents).
* Stay current with best practices in AI safety, compliance, and governance (outsourced partners like Vanta where relevant).
* Contribute to technical architecture, product decisions, and early client pilots.
Qualifications
* Strong experience with LLM frameworks (OpenAI API, LangChain, LlamaIndex, Hugging Face).
* Proficiency in Python and TypeScript/JavaScript (for full-stack or backend integration).
* Solid understanding of prompt engineering, RAG, and agentic workflows.
* Familiarity with evaluation frameworks (Evals, Weights & Biases, Trulens, LangSmith).
* Understanding of reinforcement learning concepts — especially RLHF / RLAIF and how to implement continuous improvement loops.
* Knowledge of vector databases, embeddings, and data ingestion pipelines.
* Experience deploying applications on major cloud providers (AWS, Azure, GCP) and working with developer platforms like Supabase or Vercel.
* Startup mindset: hands-on, fast-moving, product-driven.
Bonus:
* Background in behavioural science, psychometrics, or HR tech.
* Experience in PE, VC, or executive assessment domains.
* Prior experience as a founding or early engineer.