Understand the tool
Students learn what the system is doing
Not just which button to press. The aim is better judgement about where agentic AI helps, where it misleads, and what still needs to be done independently.
What students leave with
Clear judgement, better habits, and useful systems.
Use it properly
AI supports the thinking instead of replacing it
The line is clear: use AI to plan, explain, critique, organise, and test understanding. Do not outsource the actual thinking that school and university work is supposed to build.
Build a system
Every session ends with something reusable
A prompt pattern, a note structure, a revision loop, a workflow map, or a personal AI use policy that the student can keep using after the session ends.
How it works
Simple format. Personalised to the student.
Open materials
See the tutorial site and workshop series already available.
Open-source tutorial
Agentic AI in life sciences tutorial
A public tutorial site built from the same teaching approach: practical workflows, real tools, and material people can use on their own.
Workshop playlist
Watch the talks and workshop walkthroughs
The video series shows the teaching style directly, including the life sciences introduction and the wider workshop playlist.
The programme
Six steps. One clear progression.
0
Intro call and meeting
A free first conversation to understand the student, their subjects, their current AI use, and what the programme should focus on.
1
Introduction to agentic AI tools
Claude Code, Codex, OpenCode, and Copilot. What each tool is for, where it helps most, and how to choose between them.
2
Context engineering
How to structure instructions, source material, and constraints so the model gives better answers.
3
RAG, skills, and agents
How retrieval, reusable workflows, and agent systems fit together in practice.
4
Building a second brain
How to organise notes, readings, and project material in systems like Obsidian.
5
Teaching yourself with AI
How to use AI to explain, quiz, critique, and plan your own learning.
Built from real work
Research, bots, hackathons, and hands-on teaching.
Open teaching, practical workflows, real examples.
The same approach works in one-to-one tutoring, small-group sessions, hackathons, and research workshops.
115+
Registered attendees for the ClawBio hackathon teaching at Imperial College London.
4
Imperial Agentic AI workshops in the talk series and related teaching sessions.
10
Researchers trained directly in a lab-group setting on hands-on AI workflows.
Live
Bots and agent environments already deployed for research, operations, and genomics workflows.
Products and systems
ClawBio, RoboTerri, UK DRI Bot, Lab-Agents
Built tools for Telegram and Slack covering genomics, lab workflows, literature, grants, institutional knowledge, and agentic research support.
Teaching formats
Hackathons, workshops, small groups
From ClawBio hackathon teaching to UK DRI and Imperial workshops, plus the open-source Agentic Life Sciences Tutorial, the style stays practical, tool-first, and grounded in how the systems actually work.
Built for beginners
Designed to help people start well
The same approach has been used to teach research managers, students, and researchers: reduce the jargon, keep the workflow real, and build confidence through use.
About Jay
Research-led, practical, and built around real student work.
I’m a PhD researcher at Imperial College London working in computational biology and AI at the UK Dementia Research Institute. Alongside research, I build agentic tools, run workshops, teach hackathon sessions, and help people adopt these systems in ways that are actually useful.
The aim here is not to make students dependent on AI. It is to help them understand the tools, use them appropriately, and build better workflows for learning, organisation, writing, research, and revision.
Get started
Plan a first session around the student’s real work.
The first meet and greet is free. We can use it to map the student’s subjects, goals, AI experience, and whether the programme should lean more toward writing, revision, organisation, coding, or research.