Robotics / AI / Systems

Robots,
intelligent systems,
and the infrastructure
between them.

Selected work across open robotics, physical AI, healthcare platforms, and retrieval systems—built by Alok Didde.

A softly animated industrial robotic arm
Alok soldering electronics beside a 3D printer in his workshop

01 / Selected work

Things made,
tested, and learned from.

Systems are most interesting at their edges: where software meets hardware, models meet messy data, and prototypes meet actual use.

Animated diagram of a modular robot system
Fig. 01
Robotics infrastructureBuilding now

An open stack for building real robots

A framework for moving from an idea to working hardware without rebuilding the same infrastructure every time.

The aim is practical: composable software, firmware, and hardware interfaces that are understandable enough to change—and solid enough to build on.

  • ROS 2
  • C++
  • Python
  • Open source
Animated diagram of an autonomous agricultural robot
Fig. 02
MSc Robotics projectBuilt 2024

A field robot that sees, navigates, and acts

An autonomous agricultural platform combining spatial reconstruction, visual pest detection, and robotic-arm control.

Gaussian Splatting informed navigation, CNNs handled pest detection, and forward/inverse kinematics drove precise manipulation.

  • Gaussian Splatting
  • Computer vision
  • Kinematics
Animated diagram of exploratory information retrieval
Fig. 03
Retrieval researchExperimental

Retrieval that explores instead of only ranking

A retrieval approach that uses Monte Carlo exploration to search a corpus rather than depending entirely on vector similarity.

In early experiments it reached roughly 80% recall on the target task, compared with under 20% for the embedding baseline. The interesting part is not the number—it is the search behaviour behind it.

  • RAG
  • Monte Carlo search
  • LLMs
Animated technical diagram of a circuit board designed as code
Fig. 04
Agent-native EDAOpen source

Circuit boards as code, with evidence

PCBoo lets coding agents author circuit boards as composable TypeScript, then makes them show their working.

Deterministic builds, structured diagnostics, local inspection, and separate readiness statuses keep a plausible-looking board from quietly becoming a very expensive coaster.

  • TypeScript
  • Bun
  • tscircuit
  • Hardware
Read the project story
Animated diagram of connected healthcare systems
Fig. 05
Software systems14 years in practice

Healthcare systems designed for the messy real world

Clinical platforms, remote patient monitoring, real-time communication, claims workflows, and AI-assisted document systems.

Work across startups and scaled healthcare companies shaped how I think about reliability: the difficult part is rarely the happy path.

  • Distributed systems
  • Healthcare
  • Applied AI

02 / Working questions

The thoughts behind the work.

Not declarations. Questions worth keeping open long enough to build something around them.

A

How much of robotics infrastructure can become boring?

The useful kind of boring: predictable, inspectable, and reusable.

B

What changes when retrieval explores instead of ranks?

Search may need strategies, memory, and backtracking—not only a better embedding.

C

What makes open hardware genuinely extensible?

Documentation and interfaces matter as much as a downloadable CAD file.

A short signal

Fourteen years building software. Now applying that scar tissue to robots.

Focus
Open robotics + physical AI
Background
Healthcare and distributed systems
Training
MSc Robotics, Queen Mary
Based
United Kingdom

04 / Contact

Building something at the edge of software and the physical world?

Let’s talk ↗