Agentic AI & Autonomy Tooling
Give a model real tools, and be deliberate about which ones.
Most AI projects stall in the same place: the model is fine, and nobody has decided what it is allowed to touch. I build the layer in between — typed tool and resource boundaries, usually over MCP — so an assistant can operate real systems without being handed something loaded. That runs through to putting models inside existing engineering workflows, and to the case where inference has to happen on the vehicle itself, inside a power and latency budget.
What that usually looks like
- 01Wrap an existing system — hardware, database, internal service — in an MCP server an assistant can drive safely
- 02Design the tool and resource boundary for an agent that touches production or physical systems
- 03Put a model inside an existing engineering workflow where it earns its place
- 04Get a model running on a target board within a power and latency budget
- 05Build an on-device inference pipeline with sensor input and real-time output
- 06Work out whether an agentic or edge deployment is feasible at all
Where this shows up
Reconfigurable CubeSat Clusters
A compute and communications stack that turns a swarm of 1U CubeSats into a single elastic cluster, with FPGA payloads that change function in orbit and workloads that migrate between satellites in seconds.
Read →02Air-and-Water Autonomous Vehicle
AQUAD: a single vehicle that flies to a site, submerges, maneuvers underwater and returns to the air, on one charge. Built against a brief targeting ocean worlds such as Europa.
Read →03Autonomous Aerial Docking and Recharge
A heavy-lift carrier UAV acting as an airborne charging hub and data relay, letting micro-drones dock in flight to recharge, offload sensor data and borrow GPU cycles.
Read →04MCP Servers for Agentic Vehicle Control
A set of Model Context Protocol servers that give an AI assistant real, typed control of a MAVLink vehicle — arm, take off, navigate, read telemetry — plus the terrain, vision and camera context it needs to make those calls sensibly.
Read →05Natural-Language Flight Control
A chat-to-fly interface for UAV operators: spoken or typed instructions translated directly into flight commands by a language model running on the aircraft, not in a datacentre.
Read →06Real-Time Gaussian Splatting for Vehicles
A framework that streams RGB-D from a camera, drone or dataset and builds a 3D Gaussian-splat map as it goes, with a browser dashboard and an automated benchmark harness across five splatting engines.
Read →07Airborne Observatory with On-Board Plate Solving
A UAV that lifts a stabilized telescope clear of the worst of the atmosphere and works out what it is looking at on board, without needing a ground station in the loop.
Read →08Post-Quantum Telemetry for Spacecraft Links
Encrypting and signing a live vehicle telemetry link with lattice cryptography, on RISC-V compute small enough and cheap enough to fly, inside a millisecond budget.
Read →09Encrypted Command Link for a Ground/Air Robot Pair
Ascon protecting the command channel between a ground vehicle and an aircraft, with an encrypted heartbeat that fails both vehicles safe the moment the link dies.
Read →