Talks & Presentations | Rabimba Karanjai

LA Tech Week 2026 — Google for Developers Codelabs & AI Workshops

Thu, 15 Oct 2026 14:00:00 +0000

Workshop Overview

At LA Tech Week 2026, Google for Startups and Google for Developers hosted hands-on codelab tracks for enterprise founders, engineers, and technical leaders. Alongside fellow Google Developer Expert Kiruthika Subramani, I co-instructed two interactive technical workshops detailing how to build production-grade agentic architectures with Google Cloud and the Model Context Protocol (MCP).

Workshop 1: Location Intelligence ADK Agent with MCP Servers

Interactive Presentation & Teleprompter:
Official Google Codelab:
Core Technologies: Google Agent Development Kit (ADK), BigQuery Geospatial Data, Google Maps Grounding Lite MCP, Gemini 2.5 Flash

Architectural Focus

Agent Development Kit (ADK) Core: Structuring autonomous agents with schema-enforced tool interfaces, state persistence, and error-recovery loops.
BigQuery MCP Integration: Bridging spatial data warehouses with natural language queries—enabling dynamic SQL generation, bounding box filtering, and spatial joins over enterprise logistics tables without manual query plumbing.
Google Maps Grounding Lite MCP: Validating place details, computing multi-modal travel times, and grounding model answers against live physical coordinates.
Production Guardrails: Rate-limiting spatial queries, handling disconnected MCP sessions, and sanitizing agent reasoning steps.

Workshop 2: Command & Control: Gemini, MCP & Agentic CLI

Interactive Presentation & Teleprompter:
Official Google Codelab:
Core Technologies: Antigravity CLI (agy), Developer Knowledge MCP, Agent Skills, Nano Banana 2 Image Gen, Google Cloud Run

Architectural Focus

Headless Agent Orchestration: Driving terminal-first agentic developer loops with Antigravity CLI (agy), managing multi-step project scaffolding and debugging sessions.
Developer Knowledge MCP: Exposing local codebase indexing, symbol trees, and architecture specs to agent runtimes without manual context dumping.
Multimodal Generation & Deployment: Chaining multimodal synthesis via Nano Banana 2 for visual assets directly into automated container packaging and deployment to Google Cloud Run.
Composable Agent Skills: Packaging reusable capability modules across teams to enforce testing, security, and deployment standards across agent prompts.

Reception & Industry Feedback

“Google for Startups and Google for Developers take over #LATECHWEEK… Thanks to my GDEs Rabimba Karanjai and Kiruthika Subramani for providing such wonderful and thought driven codelabs.” — Marty Fortier, Google Developer Ecosystem / Startup Lead

“LA TECH WEEK by a16z was hands-on. We built small tools around real questions from our day-to-day, like which truck a load needs and how to read proof-of-delivery paperwork faster, and got a map agent running in the cloud in a guided codelab… Thank you to Rabimba Karanjai and Kiruthika Subramani, our instructors in the Google for Developers codelabs.” — Elizabeth Moscoso, Founder & Logistics Operator

House of AI: AI Safety Forum — Closing Panel

Sat, 12 Sep 2026 16:40:00 -0700

Context Contamination at USENIX Security 26 — Oral Presentation & Poster

Wed, 12 Aug 2026 09:00:00 +0000

OpenXLA Summer DevLab 2026 — Benchmarking vLLM on XLA Backends

Thu, 25 Jun 2026 18:00:00 +0000

Talk Overview

Invited by Google’s ML and TPU teams to present at OpenXLA Summer DevLab 2026 hosted at Google Corporate HQ in Mountain View, CA, alongside speakers from Google, NVIDIA, and the Meta PyTorch team.

The talk details our empirical evaluation of the matured vLLM + OpenXLA stack, contrasting it with the native vLLM + CUDA serving path under identical enterprise workloads.

Key Technical Insights

1. The Runtime Collision: Dynamic CUDA vs. Static OpenXLA

GPU / CUDA: Inherently dynamic, memory allocation and kernel launches scale on the fly to arbitrary prompt lengths with near-instantaneous container startup.
TPU / OpenXLA: Built around high-density Systolic Arrays streaming deterministic matrix multiplications. XLA compiles rigid machine code ahead of time, requiring incoming variable-length prompts to be mapped into static padding buckets (128, 256, 512, 1024 tokens), introducing compute padding overhead.
The Compilation Wall: Uncompiled sequence configurations or cold starts trigger ahead-of-time compilation passes that can pause server execution for 20–30 minutes, complicating elastic autoscaling.

2. The Throughput Trap vs. Usable Goodput

Standard evaluations optimize for Peak Throughput (tokens/sec), masking degraded latency.
LLM inference splits into:
- Prefill phase: Compute-bound; dictates Time to First Token (TTFT).
- Decode phase: Memory-bandwidth-bound; dictates Time Per Output Token (TPOT).
We evaluate capacity using Usable Goodput — requests per second successfully meeting both TTFT ≤ 1000 ms and TPOT ≤ 50 ms simultaneously.

3. Interconnect Limits & Tail Latency Collapse

On single-chip configurations (e.g., Llama-3.1 8B), the system scales predictably, keeping P99 response loops under 38 ms.
For partitioned 70B models requiring inter-chip all-reduce tensor collectives, operating over baseboard PCIe Gen5 without NVLink bridges quickly saturates the bus. P99 decode tail latency spikes to ~466 ms/token, causing usable goodput to collapse from 9.3 down to 2.4 requests/second under peak concurrency.

4. The 2026 Unified vLLM-XLA Stack Architecture

Consolidated Control Layer: Google’s internal JetStream stack was officially archived in early 2026 and folded into the JAX-native tpu-inference engine, with vLLM providing the OpenAI-compatible frontend orchestrator.
Torchax Lowering: Standard PyTorch graphs are caught on-the-fly and lowered into StableHLO structures without eager Python interpreter bottlenecks.
Ragged Paged Attention v3 (RPA v3): Implemented in Google’s low-level Pallas and Mosaic kernel layout engines, RPA v3 enables dynamic ragged tiling inside HBM, fusing KV-cache scatter steps into the attention block and delivering up to 86% Model Bandwidth Utilization (MBU) during generation.

Resources & Artifacts

Presentation Slides:
Benchmark Harness & Test Suite:
Deep-Dive Engineering Post:
Recorded Session Stream:

Google I/O GDE Field Test — AI Race Coach at Sonoma Raceway

Sat, 23 May 2026 08:00:00 +0000

Beyond the Chatbot: A Blueprint for Trustable AI

Wed, 15 Apr 2026 10:00:00 +0000

United in Defense: Architecting Safe and Trustworthy AI Agents

Fri, 20 Mar 2026 10:00:00 +0000

Google I/O 2025: GDE Summit, From Whiteboard to Users

Sun, 01 Jun 2025 10:00:00 +0000

From Whiteboard to Users: Making Research Accessible

Sun, 18 May 2025 10:00:00 +0000

Unmasking the Shadows: AI Red Teaming in the Age of Gemini and VertexAI, fortified by SAIF

Sat, 05 Apr 2025 10:00:00 +0000

Supercharging GenAI: Ray, Kubernetes, and TPUs for Lightning-Fast Inference

Fri, 15 Nov 2024 10:00:00 +0000

LLM, Reasoning and Agentic Gemma

Tue, 12 Nov 2024 10:00:00 +0000

LLMinABox: On Device Personalized Diary & Concierge using your voice and Gemma

Tue, 24 Sep 2024 10:00:00 +0000

GDG Houston Presents: IO Extended Web Edition. The state of WebXR by Rambimba Karanjai

Thu, 20 Jun 2024 10:00:00 +0000

Talk to your Desktop, Code, PDF or anything, in any language, using your own LLM or Gemini

Tue, 18 Jun 2024 10:00:00 +0000

Walkthrough of Gemini and Gemma on Cloud and On-device

Tue, 18 Jun 2024 10:00:00 +0000

Google I⧸O Extended Edmonton 2024 Immersive art with AI

Sat, 15 Jun 2024 10:00:00 +0000

Stanford CS25: V2 I Emergent Abilities and Scaling in LLMs

Wed, 12 Jun 2024 10:00:00 +0000

LangChain Deep Dive with Rabimba Karanjai (Google Developer Expert)

Mon, 10 Jun 2024 10:00:00 +0000

This is what we mean when we say Google Developer Experts (GDE)

Mon, 10 Jun 2024 10:00:00 +0000

TPU As Cryptographic Accelerator: At Google Cloud Research Innovator Showcase

Mon, 10 Jun 2024 10:00:00 +0000

GDG Zagreb - Online Tech Talk #1

Thu, 06 Jun 2024 10:00:00 +0000

GDG Zagreb - Online Tech Talk #1 - State of WebXR and building with AR/VR today - Rabimba Karanjai

Thu, 06 Jun 2024 10:00:00 +0000

Collaboration Is All You Need: LLM Assisted Safe Code Translation

Wed, 05 Jun 2024 10:00:00 +0000

SolMover: https://dl.acm.org/doi/10.1145/3664646.3664771

Wed, 05 Jun 2024 10:00:00 +0000

Mitigating Hallucinations in AI-Driven Medical Diagnosis

Mon, 20 May 2024 10:00:00 +0000

AI, LLMs, and the Future of the Web: A Conversation with Rabimba and Hemanth (PayPal) at Google I/O

Tue, 14 May 2024 10:00:00 +0000

Berkley RDI: Code Translation

Wed, 10 Apr 2024 10:00:00 +0000

SolMover: Smart Contract Code Translation Based on Concepts

Wed, 10 Apr 2024 10:00:00 +0000

Build with AI Intro to Function Calling with Gemini by Rabimba Karanjai GDE 2024 03 20 18 07 GMT 5

Wed, 20 Mar 2024 10:00:00 +0000

Hands-on Workshop: Develop an App with Vertex AI Gemini 1.0 Pro

Wed, 20 Mar 2024 10:00:00 +0000

Privacy aware Zero Knowledge Login using Oauth2 and Passkey

Wed, 20 Mar 2024 10:00:00 +0000

Give your web apps superpower with Generative AI and Mediapipe

Wed, 28 Feb 2024 10:00:00 +0000

LLM Applications components and design patterns Hands on Workshop

Tue, 20 Feb 2024 10:00:00 +0000

On Device Generative AI: Building your own Dall-E in the browser, welcome WebGPU

Wed, 08 Nov 2023 10:00:00 +0000

An Empirical Study of AI-based Smart Contract Creation

Thu, 02 Nov 2023 10:00:00 +0000

DeepSpeech: A Journey to <10% Word Error Rate

Fri, 22 Sep 2023 10:00:00 +0000

North America Connect 2022 ｜ Day 2 GDE Stage Rabimba

Sun, 04 Dec 2022 10:00:00 +0000

DeepSpeech Architecture at North Carolina DevFest 2022 at Chapel Hill

Wed, 30 Nov 2022 10:00:00 +0000

State of WebXR: What do You Need to Know Today? - Rabimba Karanjai - NDC Sydney 2020

Tue, 01 Sep 2020 10:00:00 +0000

Building Decentralized Social Virtual Reality using WebXR on your browser

Sun, 02 Feb 2020 10:00:00 +0000

FOSDEM 2020 - Building Decentralized Social Virtual Reality using WebXR on your browser

Sun, 02 Feb 2020 10:00:00 +0000

Testing Web VR/AR Mixed Reality Applications: What you need to know by Rabimba Karanjai #SeConf2020

Wed, 15 Jan 2020 10:00:00 +0000

Mozilla TS Paris Meetup Talk : Introducing jARvis

Thu, 09 Jan 2020 10:00:00 +0000

Voice Driven VR

Mon, 04 Nov 2019 10:00:00 +0000

DevFest Ranchi 2019 – Rabimba Karanjai - DeepSpeech: A Journey to less than 10% Word Error Rate

Fri, 04 Oct 2019 10:00:00 +0000

Data Driven Innovation

Wed, 10 Oct 2018 10:00:00 +0000

DevDay Nang

Thu, 20 Sep 2018 10:00:00 +0000

University of Texas at Dallas Welcome Week Tech Talk

Wed, 22 Aug 2018 10:00:00 +0000

O'Reilly OSCON 2018 Talk

Wed, 18 Jul 2018 10:00:00 +0000

O'Reilly OSCON 2018 Tutorial

Tue, 17 Jul 2018 10:00:00 +0000

RevolutionConf

Thu, 07 Jun 2018 10:00:00 +0000

Visualize your Data in a 3D VR world using A-Frame in WebVR

Sun, 20 May 2018 10:00:00 +0000

Fullstack London

Wed, 16 May 2018 10:00:00 +0000

State of WebVR, A-Frame and Our Foray Into Web Mixed Reality : Rabimba Karanjai - Mozilla

Thu, 19 Apr 2018 10:00:00 +0000

DevWeek NYC

Thu, 15 Mar 2018 10:00:00 +0000

DeveloperWeek Austin

Wed, 07 Feb 2018 10:00:00 +0000

Mexico City Mobile Day

Wed, 15 Nov 2017 10:00:00 +0000

Rabimba Karanjai: VR for Humans Build your world using WebVR and aframe | Mobile day 2017

Wed, 15 Nov 2017 10:00:00 +0000

Google Chrome Dev Summit

Wed, 01 Nov 2017 10:00:00 +0000

Turning sensors into signals

Sun, 15 Oct 2017 10:00:00 +0000

Hardening Your IoT Endpoints: A Preventive Toolkit

Mon, 25 Sep 2017 10:00:00 +0000

LinuxCon 2017

Mon, 25 Sep 2017 10:00:00 +0000

Singapore Leadership Summit: Connected Devices and Mozilla TechSpeaker

Wed, 20 Sep 2017 10:00:00 +0000

Kansas City Developer Conference 2017

Wed, 28 Jun 2017 10:00:00 +0000

Open Source Bridge 2017

Wed, 21 Jun 2017 10:00:00 +0000

Optimizing Web Virtual Reality

Thu, 15 Jun 2017 10:00:00 +0000

MozLondon: Connected Devices Team

Wed, 14 Jun 2017 10:00:00 +0000

ACM SIGGRAPH Web3D

Wed, 07 Jun 2017 10:00:00 +0000

SecurityPI: IronClad your Raspberry PI

Mon, 15 May 2017 10:00:00 +0000

State of WebVR & aframe: Yesterday, Today, and Tomorrow Beyond Horizon

Thu, 20 Apr 2017 10:00:00 +0000

Silicon Valley Virtual Reality

Sat, 15 Apr 2017 10:00:00 +0000

LibrePlanet 2017

Sat, 25 Mar 2017 10:00:00 +0000

Securi-Tay 2017

Fri, 10 Mar 2017 10:00:00 +0000

DeveloperWeek 2017

Wed, 15 Feb 2017 10:00:00 +0000

Linux Foundation OpenIoT Summit 2017

Wed, 15 Feb 2017 10:00:00 +0000

Building Your World in WebVR

Mon, 13 Feb 2017 10:00:00 +0000

Are you well? When connected devices know more about you

Sun, 30 Oct 2016 10:00:00 +0000

Painting with Virtual Reality and A-Painter

Sat, 29 Oct 2016 10:00:00 +0000

Build on WebVR and touch to interact! Come Augmented Reality for the web!

Fri, 28 Oct 2016 10:00:00 +0000

GraphicalWeb 2016

Wed, 12 Oct 2016 10:00:00 +0000

ViewSource Berlin (Game Demo and VR Booth)

Thu, 15 Sep 2016 10:00:00 +0000

MozTechSpeaker Berlin Meetup

Tue, 13 Sep 2016 10:00:00 +0000

BrazilJS VR Talk

Sat, 20 Aug 2016 10:00:00 +0000

OpenSource Bridge 2016

Wed, 22 Jun 2016 10:00:00 +0000

OpenSource HongKong

Sat, 14 Nov 2015 10:00:00 +0000

Rabimba Karanjai - A self learning word prediction module for handling multilingual input

Thu, 24 Sep 2015 10:00:00 +0000

JSFoo 2015

Fri, 18 Sep 2015 10:00:00 +0000

Open Source Bridge 2015

Wed, 24 Jun 2015 10:00:00 +0000

OSB 2015 - Building a self learning word prediction and auto-correct module for FirefoxOS

Wed, 24 Jun 2015 10:00:00 +0000

WikiConfUSA 2014

Sat, 07 Jun 2014 10:00:00 +0000

QA Robotics at Confluence 2012

Wed, 20 Jun 2012 10:00:00 +0000