Skip to content
Atlas profile

Jordan Legg

@404missinglinkVerified GitHub profile

Chief AI Officer at takara.ai | Leads Applied and Frontier AI/ML Teams

AI-Atlas score

Advanced
0100
Strongest
Machine Learning
60.8/100 domain score
Rank
39th=
of 86
United Kingdom
2nd
of 3
  • United KingdomLocation
  • 15 Nov 2026Mapped
  • 6 / 15AI projects
01Skills matrix

AI Skills Matrix

AI Skills Matrix scores, 0 to 100
Machine Learning61
Computer Vision36
Data Science17
NLP & LLMs50
AI Agents57
MLOps45

Each node is one domain, scored only from evidence in these public repositories. Select one to see why.

Substantial implementations with real depth.

02Evaluated projects6 of 6

15 public repositories → 6 with AI work, evaluated. The other 9 (repositories without AI code, forks and empty repositories) are not scored and have no effect on the result. How repositories are selected

Ranked by contribution

Ordered by how much each project adds to the Atlas score, not by its own score. A project with evidence in several domains can rank above one with a slightly higher project score. Choose a domain to focus on the projects with meaningful evidence there.

01
56.0

Implementation

24% of Atlas Score

Educational Implementation· Python

The repository implements a token memory simulation using graph diffusion, where token strengths spread over a co-occurrence graph with leakage and stimulus injection. It provides a runnable demonstration that outputs simulation statistics, plots, and metrics.

Machine Learning0.80NLP & LLMs0.70Data Science0.30
02
56.2

Implementation

20% of Atlas Score

AI Developer Tool· Python

WIIIIIDE is a distributed training system for PyTorch models that enables training across multiple nodes with a central server, featuring encrypted communication, gradient compression, and a training visualizer.

Machine Learning0.90MLOps0.40
03
55.0

Implementation

16% of Atlas Score

AI Developer Tool· Go

go-code-runner is an MCP server written in Go that provides a single tool to compile and execute user-provided Go snippets inside a containerized sandbox, using the Model Context Protocol for communication with AI agents.

AI Agents0.90
04
45.7

Implementation

15% of Atlas Score

MLOps Infrastructure· Dockerfile

The repository provides a Dockerfile that builds a container image with PyTorch, TorchCodec, and a custom FFmpeg build to ensure compatibility, along with a simple verification script.

MLOps0.80Machine Learning0.50Computer Vision0.20
05
34.6

Exploration

13% of Atlas Score

ML Application· Python

Pix4D-ArchScan is a video to text summarization tool that extracts frames from MP4 videos and uses the Mistral AI API to generate frame descriptions and an overall summary. It provides a Gradio-based user interface for uploading videos and viewing the summaries in a chat format.

Computer Vision0.80NLP & LLMs0.80AI Agents0.20
06
64.8

Advanced

11% of Atlas Score

AI Developer Tool· Go

Operator is a hybrid-search MCP gateway for Go that routes natural language to the right MCP using vector + keyword fusion, optional LLM disambiguation, and live session management. It provides an SDK, TUI chat, HTTP API, and stdio MCP server mode.

NLP & LLMs0.60AI Agents0.40