What I do

I take on hard core technical problems in applied AI, scientific computing, and research engineering, typically where the existing approach has hit a wall and the path forward isn’t obvious.

Not a vertical specialist, not a strategist, not someone you bring in to convince a stakeholder or run a team. The work is the problem itself: getting a difficult system from prototype to something that runs reliably in production, or turning a research idea into a clean implementation a team can build on.

Where I’m strongest

• Applied AI. Designing and shipping ML and generative systems for problems that don’t fit off-the-shelf tooling, including real-time inference, reinforcement learning, and small LLMs running under tight latency and compute budgets.
• Scientific computing. Numerical methods, simulation, and large-scale compute on heterogeneous hardware (CPU and GPU).
• Research engineering. Taking a paper, prototype, or internal research effort and producing a validated, well-engineered implementation.
• Algorithm design. Graph-theoretic, combinatorial, and partition-based algorithms, with computational backing where it matters.

How engagements tend to look

Short, focused, technically deep. Often a few weeks of investigation and prototyping that produce a working approach your team can run with, not a deck.

Availability

Currently available for a small number of new engagements when there’s a good fit. If your problem sounds like something I might help with, please reach out.

Contact

Email: david@lixel.io