> Markdown version of [/jobs/ext/3882762-graduate-ai-engineer-pe-firm-cto-office](https://www.wearedevelopers.com/jobs/ext/3882762-graduate-ai-engineer-pe-firm-cto-office). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Graduate AI Engineer- PE Firm-CTO office - **Company:** Robert Walters - **Location:** London, United Kingdom - **Experience:** Starter - **Contract:** Permanent contract - **Skills:** JavaScript (Programming Language), Application Programming Interfaces (APIs), Artificial Intelligence, Python (Programming Language), OpenAI, TypeScript, Large Language Models, Information Technology, Software Coding - **Published:** October 23, 2026 - **Apply:** https://www.reed.co.uk/jobs/graduate-ai-engineer-pe-firm-cto-office/57484912 ## About the Role * A degree (2:1 or above, or equivalent evidence of ability) in computer science, engineering, maths, physics, or a similarly quantitative field or a demonstrable self-taught equivalent. * Solid programming ability in Python (and ideally some TypeScript/JavaScript for lightweight UIs). You can show us things you've built. * Hands-on curiosity about LLMs: you've built something with a model API (OpenAI, Anthropic, open-weights), even if it's a personal project. * Product sense: you care whether people actually use what you build, and you'd rather ship something rough this week than something perfect next quarter. * Clear communication. You'll demo to and train non-technical investment professionals. * Care and discretion. You'll handle commercially sensitive financial data and must take security and confidentiality seriously. ## Description * Build document intelligence. LLM-based extraction of structured data from the documents private markets run on: CIMs, financial statements, term sheets, legal agreements, DDQs. * Build the knowledge layer. Retrieval-augmented search over the firm's deal history, research, and documents, so institutional knowledge stops living in inboxes. * Automate workflows. Deal screening and market scanning, investor reporting, DDQ/RFP response drafting - anything repetitive that specialists currently do by hand. * Own evaluation. Build the harnesses that measure whether AI outputs are accurate enough for an investment context. Precision matters enormously here, and "looks plausible" is not a bar. * Drive adoption. Train the deal team on the tools you ship, gather feedback, run internal demos and lunch-and-learns, and iterate fast on what you learn. * Prototype relentlessly. Evaluate new models, frameworks, and vendor tools; build quick proofs-of-concept; recommend build-vs-buy with evidence.