Why Forward-Deployed Engineers Are the Secret Weapon for AI in Legacy Industries

Earlier this year, our CTO and Co-Founder, Hunter Honnessy, joined Alex Gras from Mercury Fund and John Kalfayan from Collide for a panel at Mercury Fund Day at the Ion in Houston. They discussed the New FDE Model and how to deploy it at scale in legacy industries.

This conversation has been growing across the AI world, and it feels especially relevant to us at Unlimited.ai. We want to share some key themes from the discussion and explain why we believe the FDE model is central to how AI gets adopted in industries like private markets.

Check out the video link below, and read through our deeper dive below that.

First, What's an FDE?

FDE means Forward Deployed Engineer. Companies like Palantir started the idea, but now it’s become common. In 2025, job postings for FDE roles jumped by over 800%, and companies such as Salesforce and OpenAI now have their own FDE teams.

The idea is straightforward. Instead of just sending software and hoping customers can use it, engineers work directly with the customer. These engineers are the builders who understand both the technology and the customer’s daily challenges, such as messy data, old workflows, and half-documented systems that every business faces.

This is especially important for AI products. A model might look impressive in a demo but be useless in real situations if no one helps connect what the AI can do with what the business actually needs.

Why This Matters for Private Markets

At Unlimited.ai, we’re creating AI-based portfolio management for family offices and RIAs. These investors handle complex portfolios across private equity, venture capital, real estate, and more, often with small teams overwhelmed by PDFs, spreadsheets, and emails from multiple fund managers.

This is the kind of situation where you can’t just give someone a login and wish them luck. The data is messy, the workflows are set in their ways, and every family office works a bit differently.

That’s why the FDE model fits so well with what we’re building. When you deploy AI agents that handle sensitive financial data, like parsing LPAs, reconciling capital calls, or showing performance insights, you need people who can work directly with customers and make sure the system works for them, not just in a demo.

The Feedback Loop Is the Real Advantage

One of the most important parts of the FDE model is not just what gets delivered to the customer, but also what comes back. Every unusual case an FDE finds, every workflow detail they notice, and every piece of feedback they gather helps improve the product for everyone.

This is especially true in older industries like private markets, where the real complexity comes from the business itself, not just the algorithms. For example, a $500M family office manages capital calls differently than a $5B multi-family office. An FDE working directly with clients can understand these differences in a way that a product team in a conference room cannot.

Agentic AI Needs a Human Layer

The panel also discussed how agentic AI and blockchain come together, which are two main investment themes for Mercury Fund. As AI agents become more independent and handle tasks like document retrieval, due diligence, and portfolio reporting without people constantly watching over them, trust becomes even more important.

How do you make sure an AI agent is doing the right thing when it’s working with someone’s most sensitive financial data? You need a human layer in the process—engineers who can set up, monitor, and adjust these systems in real situations. That’s what FDEs do. They connect powerful AI tools to real business results.

Building for the Long Game

The discussion at Mercury Fund Day confirmed something we strongly believe at Unlimited.ai: the companies that succeed in enterprise AI will not just have the best models. They will also be the best at deploying them. These companies will show up, work closely with their customers, and do the hard work to make AI work in places that were not designed for it.

We appreciate Mercury Fund for hosting the conversation, and thank Alex and John for a great discussion.

If you’re a family office or RIA interested in seeing how AI-powered private markets orchestration works in practice, we’d be happy to show you. Please reach out.