AI Strategist

AI Strategist

BrainLake | London | Full-time

About BrainLake

BrainLake transforms private alternative asset managers into AI native funds. We succeed by providing funds with tailored AI training programmes, a continuous learning platform and building custom workflows and agents for alternative asset professionals in private equity, credit, real estate, secondaries, fund of funds, and family offices.

Fund teams are already using AI, but most are only scraping the surface of what's possible: ad-hoc ChatGPT prompts, no structure, no measurement, no connection to their actual workflows. We entrench AI efficiency throughout their workflows through custom development and education. Our structured education includes hands-on training days, a learning platform with scored exercises and curated workflow libraries.

We have a deep background in building agents for financial professionals and deep sector experience across Goldman Sachs, Credit Suisse, and Target Global. This is an early-stage hire. You will be building alongside the founders, not joining a machine that already runs.

The Role

Day to day, the job breaks down to three main components:

1. Research and test AI tools relevant to fund workflows: You will be constantly evaluating LLMs, agent frameworks, MCP integrations, and specialist tools and methodologies to maintain a practical knowledge base of what actually works for alternative asset professionals. You’ll help update the curriculum, the solutions library, and the platform content based on your findings.

2. Build custom workflows and agents for client teams: Y ou will design, build,and deliver AI workflows and agents tailored to client-specific processes. This is application-layer building using all the tools at your disposal to build the most efficient outcome for clients. Prior examples include deal screening agents, IC memo drafting agents, portfolio KPI monitoring agents, and LP reporting workflows. You scope the work via short discovery sprints, prototype with the client team, deploy into their environment, and iterate based on real usage. The point is to operationalise what the client needs.

3. Shape and deliver training: You help build and run the one-day AI Mastery Bootcamp (covering LLM fundamentals, prompting, finance-specific plugins, MCP connectors, agent building). You’ll alsocontribute to the ongoing platform courses and develop the extended programme. You are not just assembling slides. You are designing exercises that use real financial documents, building demo workflows, and writing the curriculum that makes fund professionals competent with AI tools.

What We Are Looking ForYou should be

• Genuinely curious about AI tools and how they evolve week to week. Not theoretically interested, you should be someone who tries things, breaks things, and forms first-hand opinions about what works.
• A clear communicator who can present technical concepts to a room of investment professionals with mixed levels of technical understanding. You do not need to be a polished conference speaker, but you do need to be credible and composed in front of senior fund professionals.
• Comfortable with ambiguity. Processes do not exist yet. You will help build them.
• Self-directed. You will report to the CEO, but you will not be managed hour by hour. You identify what needs doing independently, and do it.
• Organised enough to manage multiple workstreams: research, content creation, client delivery, and community building.

You should have

• Strong working knowledge of current AI tools: LLMs (Claude, ChatGPT, Gemini, Perplexity at minimum), prompting techniques, and a practical understanding of how they perform across different tasks.
• The ability to design and deliver training sessions. This could come from teaching, consulting, workshops, bootcamps, or similar - the format matters less than evidence you can hold a room and transfer knowledge effectively.
• Enough comfort with frontier models to build agents, automate workflows, demo tools live, troubleshoot in real time, and guide non-technical users through setup and configuration.

Preferred but not required

• Experience in finance, consulting, or alternative assets. You will be working with alternative asset investors, fund operations teams and Investor relations teams. Domain familiarity helps, but we can teach the finance context if you bring everything else.
• Technical skills: Python, API integrations, Git. The curriculum includes sessions on Claude Code, agent building, and custom skill creation. If you can do these things yourself, you will be a better trainer. But this is a nice to have, not a gatekeeper.
• Experience at an early-stage company or in a role where you built something from scratch.
• An interest in community building, organising events (in-person and online), creating content, building a network. This is important both for recruiting pipeline and for client acquisition.

What You Will Not Be Doing

• Building infrastructure or core platform code. The role is application-layer build using the latest models, tools and workflows. Our engineering team handles the BrainLake platform, infrastructure and software engineering needs.
• Data science or ML research. We use frontier models via API. you will not need to train models.

Working Arrangement

London-based. In-office is ideal, but we can be flexible. You will travel to client sites for training delivery (primarily UK, expanding to DACH, Benelux, and France).

Why This Might Be Interesting

You would be an early hire at a company building something mid-market funds are requesting en masse. The AI training and custom build market for alternative assets is wide open and growing.

You would work directly with the founders from day one. Your fingerprints will be on the product, the curriculum, the client relationships, and the go-to-market approach. When BrainLake grows, this role grows with it.

If you need a brand name, a large team around you, and a well-defined career ladder on day one, this is not the right fit. If you want to build something in a space where AI and finance intersect, with real autonomy and direct impact, it might be.