Tanmay
BakshiBook Tanmay

Video

Watch him hold a room

Keynotes published by the events that booked him, then the channel where the technical work lives.

Films from the events
3
Full talks, dated
7
Technical videos, at length
27
Of teaching in the library
11h

Full talks

Every talk with a recording

Each one dated, each one opening its own page.

  1. People Summit, 2026 · Keynote

    Governing the second workforce

    Who is accountable when an AI agent makes a mistake, and what changes when the systems start acting on their own.

    Watch it

  2. Google DevFest Waterloo, 2024 · Keynote

    AI is UX

    What building conversational agents for the McDonald's drive-thru at IBM taught him about users, and why AI's killer application has yet to emerge.

    Watch it

  3. INFOCOM Hyderabad, 2026 · Masterclass

    Masterclass

    A working session on what actually shifted in AI, what is hype, and how a leadership team tells the two apart.

    Watch it

  4. TED@IBM, 2017 · TED talk

    Technology that tackles the teen suicide epidemic

    An early warning system for young people at risk, and the argument that AI is worth building when it is pointed at a human problem.

    Watch it

  5. Open Source Summit North America, 2017 · Keynote

    Open sourced inspiration: the present and future of tech

    A Linux Foundation main stage session on where the tools were heading and who gets to build with them.

    Watch it

  6. nasscom Big Data and Analytics Summit, 2017 · Keynote

    AI raises the ceiling

    Opening the nasscom main stage: why AI adds to what people can do instead of replacing them.

    Watch it

  7. TEDxCincinnati, 2017 · TED talk

    Connecting the disconnected with AI

    Building assistive technology for people the mainstream tools had skipped, and what that taught him about who software is for.

    Watch it

The teaching library

27 from his own channel

Model internals and real code, from eleven minutes on what machine learning is to an hour and twenty inside JAX.

  • Finding Timestamps in Audio Transcripts by Probing IBM Granite Speech

  • Hierarchical Reasoning Model - Next-Gen Neural Problem Solving

  • Why does OpenAI have the best LLMs?

  • Can ChatGPT Replace Me?

  • How do Transformers work? Self-Attention Explained

  • BERT Masked Language Modelling: Should I pre-train on domain-specific text?

  • Getting Started with JAX: Training a Handwriting Synthesis Transformer

  • Intro to (AI?) Machine Learning

  • Perceptrons & Gradient Descent - Learn [AI] Machine Learning from Scratch Ep. 2

  • Gradient Descent in TensorFlow - Learn [AI] Machine Learning from Scratch

  • Dimensionality Reduction with Gradient Descent - Learn [AI] Machine Learning from Scratch

  • Writing K-means clustering in Python from Scratch!

  • Single-label Data for Multi-label Image Classification in TensorFlow

  • GAN Facial Super Resolution System (PULSE)

  • AlphaZero mastering the game of 2048

  • Teaching Robots to Listen with Machine Learning: Giving Pepper the Ability to Hear

  • CUDA Multi-GPU Segmented Sieve of Eratosthenes to find the Nth prime

  • NVIDIA CUDA GPU programming - processing 8 BILLION hashes/sec on V100s (Proof of Work)

  • Emscripten WebAssembly vs JavaScript: How it works & Prime Number Sieve with bit arrays

  • Write Your First LLVM FunctionPass for Instrumentation

  • CPython - Hacking in your own functionality

  • LinkedLists in Rust - Extending API Contracts to Cover Data Lifetimes

  • Async/Await: Learn Swift's Magical New Concurrency Features! WWDC 2021

  • Sudoku Playing Bot w/ Backtracking!

  • Does Competitive Programming Make You a BETTER Programmer?

  • Minecraft 1.17 + Microsoft Login NATIVE on Apple Silicon (M1, M1 Pro, M1 Max)!

  • How to Run Minecraft on Apple Silicon WITHOUT ROSETTA - Natively!

All 284 on YouTube

Send your date.

Tanmay replies himself, availability first.

Book Tanmay