All my conference / community event talks
I have been an active community speaker for over a decade. Everything is collected here, newest first, linked to slides, recordings or the event itself where those exist. The recent sessions are mostly about agentic engineering, Gemini CLI and what AI is doing to the way software gets built; the earlier ones are from a career spent on the web platform and performance.
If you would like me to speak at your conference or run a session for your team, write me a short note.
How agentic systems actually work, what it takes to build an agent, and where the practice holds up once it meets real work. For people already building with agents and for people trying to see where this goes next.
Building loops with Antigravity, and handing them the most tedious part of the work.
Harness engineering in Gemini CLI, through live demos and real workflows rather than slides — what it looks like when AI sits inside the development process instead of beside it.
A workshop at Arbisoft on putting Gemini CLI to work on real engineering tasks, with a room that pushed hard on where it holds and where it does not.
An hour with founders and students on building a Gemini CLI workspace of their own: curating organisational context, then progressive disclosure, agents, skills and hooks as the guardrails that let the system run on its own.
TEDxPunjabUniversity 2026, on the theme “Carpe Diem: Seize the Day”. One of six speakers, alongside researchers, medical doctors and a wildlife conservationist.
Three hours cutting through the noise on what actually works in harness engineering — skills, hooks, progressive disclosure, and the feedback loops that make the output improve rather than merely arrive faster.
AI Seekho 2026: Vibe Coding with Real Deployment Workflows · 200 attendedVibe coding taken past the demo and through to a real deployment workflow, with a deep dive into AI Studio and context engineering.
A hands-on workshop: agentic coding in Gemini CLI, the patterns that go with it, then applying them to mini-projects rather than reading about them.
Abstraction layers, agent orchestration, context engineering, and what happens to the junior developer role — with the argument that human judgement, architectural thinking and fundamentals matter more now, not less.
Mastering Advanced Context: Engineering For Your AI-Driven Coding Karachi, Pakistan · · 120 attendedEngineering context for precision, reliability and scale: structuring prompts and memory, and domain-specific context strategies, rather than hoping a longer prompt will do it.
Turning design sketches into working Gemini-powered prototypes, ready for real testing and market validation, deployed on Firebase.
Sixty minutes starting with how to frame the problem, then a live run through Stitch, Firebase Studio and AI Studio.
At Pak-Austria Fachhochschule: how engineers stay relevant and learn faster in the AI era, working through Gemini, Firebase Studio and ADK.
Planning an app with Gemini, designing it in Stitch, then building it view by view in Firebase Studio — prompt engineering through to production deployment.
AI code assistants, the Model Context Protocol underneath them, and how CLI tools like Gemini CLI and Claude Code are changing what the command line is for.
Beyond the IDE - Unlocking Productivity with AI Agents and Gemini CLI Lahore, Pakistan · · 200 attendedGoogle I/O Extended Lahore: code assistants, the Model Context Protocol, and the move back out of the IDE to the command line.
Generative AI in production at Taleemabad, and the education problems it is actually pointed at.
A hands-on walk through the Gemini SDK and how to use it in a real codebase.
Code assistants, the Model Context Protocol and CLI-based agents — to the largest room of the year.
Navigating the future of Software and Engineering · 60 attendedAn hour on the shift AI is forcing through the industry, with live demos of the Gemini API, AI Studio, NotebookLM and Firebase Studio.
Two hours inside the Gemini SDK with the hackathon teams, warming up on exercises before the event proper.
Ninety minutes on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes at NUST on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes at IST on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
Ninety minutes on large language models and the change they are forcing through every field, then a hands-on Colab session through the APIs.
A hands-on session on large language models, then straight into the Gemini APIs in Colab.
BuildWithAI - Gemini 101 · 24 trainedA Colab notebook worked end to end: the basics first, then caching and function calling, then putting it together.
Shaping the Future of Learning using AI and Engineering · 50 attendedAI in education, and what a personalised and genuinely accessible learning platform takes to build.
The problems facing Pakistan's education system, and where Gemini, on-device ML and a fine-tuned Gemma actually help with them.
Twenty-five minutes on the state of education in Pakistan, and on using generative AI, on-device models and Gemma to raise learning levels in public schools.
A deep dive into the Gemini API: basic generation through to multimodal input and caching.
From what learning actually is, through how neural networks borrow the same idea, up to how large language models work — and why their future is on-device.
Twenty minutes on the latest in Gemini and Gemma, from the basics through to running on-device.
A hands-on pass through the Gemini API in a Colab workbook.
Two hours, from what a large language model is through to real work in the Gemini API.
Two hands-on hours on the fundamentals of large language models.
A hands-on Gemini 101 workshop.
The basics of Gemini and prompt engineering, worked through in Google Colab.
A Gemini 101 session, part of the Build with AI series.
A hands-on session for GIKI students visiting Taleemabad.
Gen AI Workshop · 30 trainedA hands-on generative AI workshop.
A deep dive into Google AI Studio.
Building a new generation of applications on Google AI Studio and Gemini.
Mapping books to a curriculum using PaLM embeddings.
An introduction to embeddings, then a real problem: mapping books to a curriculum with the PaLM 2 API.
- Fireside: Why does every developer need to know Javascipt. Remotebase Remotefest Developer Conference, Online ·