# Jigyasa Grover

> Official digital home of Jigyasa Grover (pronounced jig-YAH-sah GROH-ver; Google Knowledge Graph: https://g.co/kgs/Lvk5yk) — Machine Learning Engineer at Uber (Tech Lead — Rider Personalization & Ranking), 13x AI & Open Source Award Winner, Google Developer Advisory Board (gDAB) member, Google Developer Expert (AI/ML), LinkedIn Learning Instructor (44,000+ learners), international keynote speaker (Google I/O main stage, 200+ talks across 20+ countries), and co-author of *Sculpting Data for ML*.

For complete structured biographical, career, speaking, course, publication, and project context, see [llms-full.txt](https://jigyasa-grover.github.io/llms-full.txt) and machine-readable JSON at [agent-context.json](https://jigyasa-grover.github.io/agent-context.json) (also exposed in-browser via W3C WebMCP `document.modelContext`).

## Primary Pages

- [Home](https://jigyasa-grover.github.io/): Overview of engineering work, speaking, courses, publications, and collaboration pillars.
- [About Jigyasa Grover](https://jigyasa-grover.github.io/about): Authoritative biography, Google & governance board roles, and Entity FAQ.
- [Experience & Career Timeline](https://jigyasa-grover.github.io/experience/): Full career at Uber, Afterdays (2015–Present), Bordo AI, Faire, Twitter/X, Meta, education (UC San Diego M.S. CS; Cornell, Maryland, Max Planck Pre-doctoral Research School; DTU B.Tech.), and 2017–2026 milestone archive.
- [Work With Me](https://jigyasa-grover.github.io/work-with-me/): Six collaboration pillars, booking links (Intro, Topmate, Passionfroot, Naano), and direct inquiry form.
- [AI Startup & Product Advisory](https://jigyasa-grover.github.io/advisory/): Independent AI advisory via Afterdays across 30+ organizations on production ML ranking, Agentic AI architecture, LLM evaluation, and token cost guardrails.
- [Keynotes & Conference Speaking](https://jigyasa-grover.github.io/speaking/): Signature talk abstracts, Google I/O keynote feature, and 200+ global appearances.
- [Corporate AI Training & Courses](https://jigyasa-grover.github.io/teaching/): Enterprise AI training & workshops delivered with LinkedIn Learning, Morgan Stanley, FedEx, Accenture, Roblox, Silicon Valley Security Group, AI Aspire, Udacity, and CommonSpirit Health, plus 4 LinkedIn Learning courses (44,000+ learners, 4.6–4.8★ across 180+ ratings).
- [Technical Content & Brand Partnerships](https://jigyasa-grover.github.io/partnerships/): Engineering-led developer content, audience demographics (82% senior technical), and brand partners.
- [Selected Engineering Projects](https://jigyasa-grover.github.io/projects/): Open-source agentic AI systems, TPU fine-tuning, edge/cloud racing telemetry, and RAG benchmarks.
- [Books & Publications](https://jigyasa-grover.github.io/publications/): *Sculpting Data for ML* (ISBN 979-8585463570, 4.8★ Goodreads, 95+ citations), *Enterprise LLM Deployment Readiness* (SSRN 2026), Springer chapters, and ACM research.
- [Press, Media & Speaker Kit](https://jigyasa-grover.github.io/press/): Forbes, Business Insider, Communications of the ACM, Google for Developers Blog, Google Cloud Tech, VentureBeat, Observer, AHCJ, Grazia UK, 13+ awards ledger, copyable bios, and press photos.
- [Quick Links & Slide QR Hub](https://jigyasa-grover.github.io/links/): Unified link-in-bio, 5.0★ 1:1 booking links, upcoming stage, courses, and downloadable conference slide QR code.
- [Technical Blog](https://jigyasa-grover.github.io/blog/): 34 in-depth technical articles and essays (plus 57 self-hosted 2015–2017 open-source archive entries at `/archive/`).

## Direct Collaboration & Booking Platforms

- [Intro (1:1 Technical & Startup Consultations — 5.0★)](https://intro.co/JigyasaGrover)
- [Topmate (Mentorship, Career Strategy & Quick Calls — 5.0★, 126+ bookings)](https://topmate.io/jigyasagrover)
- [Passionfroot (Sponsored Developer Content & Brand Campaigns)](https://www.passionfroot.me/jigyasa-grover)
- [Naano (B2B & Developer Creator Collaborations)](https://naano.com/creators/jigyasa-grover)

## Technical Articles & Essays

- [TorchTPU Doctor: Agent Skill & Performance Profiler for PyTorch on Cloud TPUs 🩺](https://jigyasa-grover.github.io/TorchTPUDoctor/) (2026-11-02): Catching silent CPU fallbacks, hidden Hugging Face host syncs, and XLA recompiles in PyTorch on Cloud TPU v6e—cutting LoRA step time 2,537× (47.3s → 18.6ms) alongside upstream PRs to transformers & peft.
- [Proof of Work - Agentic Video Understanding with Gemini as Your AI Teaching Assistant 📝](https://jigyasa-grover.github.io/ProofOfWork/) (2026-09-08): How Proof of Work uses Gemini's agentic video understanding to evaluate software engineering process, debugging discipline, and hypothesis iteration from continuous screen recordings.
- [Startup on a Shoestring - Building with a 50¢ Budget using Google Antigravity Cost & Token Monitors 💸](https://jigyasa-grover.github.io/StartupOnAShoestring/) (2026-06-28): Preventing quadratic token inflation and runaway self-correction loops in autonomous agents using Google Antigravity SDK lifecycle hooks and FastAPI SSE telemetry.
- [In my Kinetic era - Fine-tuning Gemma 3 to speak Gen Z on a Cloud TPU with one decorator 🤌🏻](https://jigyasa-grover.github.io/KineticFinetuningOnCloudTPU/) (2026-03-30): Supervised fine-tuning (SFT) of Google's Gemma 3 1B on a Cloud TPU v5 Lite using a single Python decorator—zero Docker or Kubernetes YAML required.
- [Trust at Top Speed - AI Racing Coach powered by Gemini 🏁](https://jigyasa-grover.github.io/TrustAtTopSpeed/) (2025-12-18): Building and field-testing a split-brain AI Racing Coach at Thunderhill Raceway Park combining sub-15ms on-device Gemini Nano cues, physics safety gates, and Cloud Run telemetry reasoning.
- [Supercharging VS Code Vibe Coding with Gemini Code Assist +*new* Google Colab Extension 🔋](https://jigyasa-grover.github.io/ColabVSCodeExtension/) (2025-11-14): Running modular, version-controlled Python ML workloads on cloud GPUs and TPUs directly inside VS Code using the Google Colab extension and Gemini Code Assist.
- [Critic’s Cut 🎥 Powering Multilingual, Context-Aware Movie Reviews with ADK & Gemini 2.0](https://jigyasa-grover.github.io/CriticsCutMultilingualContextADKAgent/) (2025-06-04): Architectural deep dive into Critic's Cut: building a context-aware, multilingual multi-agent system with Google's Agent Development Kit (ADK), tool integration, and PII safety callbacks.
- [Travel in Style, Rain or Shine 🧳 Smart Suitcase Packing with Gemini 2.0, Imagen 3, and Open Weather!](https://jigyasa-grover.github.io/SmartSuitcasePackingUsingGeminiImagenOpenWeather/) (2025-02-25): Combining Gemini 2.0 structured reasoning, Imagen 3 visual generation, and real-time OpenWeather API data for weather-aware travel planning.
- [Smarter than a 5th Grader? 🎓 Using LLM Comparator on Google Vertex AI to find out.](https://jigyasa-grover.github.io/SmarterThanA5thGrader-LLMComparatorOnGoogleVertexAI/) (2024-09-04): Quantitative side-by-side LLM evaluation on Google Vertex AI using PAIR's open-source LLM Comparator across accuracy, clarity, safety, and communication rubrics.
- [Book Buddy - Interactive Storytelling using a Gemma-based RAG 📖](https://jigyasa-grover.github.io/BookBuddyInteractiveStorytellingUsingAGemmaBasedRAG/) (2024-07-09): Designing a Retrieval-Augmented Generation (RAG) pipeline with Google's open-weights Gemma model for grounded interactive narratives and educational Q&A.
- [Goodbye to News Copy Writer's Block! Supercharging Gemini Pro via Supervised Fine-Tuning 📰](https://jigyasa-grover.github.io/GoodbyeToNewsCopyWriterBlock-SupervisedFineTunedGeminiPro/) (2024-03-05): Step-by-step supervised fine-tuning of Gemini Pro for structured domain adaptation and editorial headline generation.
- [Bifocal Perspective of Compound AI Systems using Gemini 1.5 Pro 👓](https://jigyasa-grover.github.io/BifocalPerspectiveOfCAS/) (2024-02-21): Examining Complex Adaptive Systems (CAS) through emergent multi-agent interactions, feedback loops, and computational modeling.
- [Hyper-Personalized Ad Campaigns using Generative AI + Quick Demo using Google Gemini Pro Vision 🎯](https://jigyasa-grover.github.io/HyperPersonalizedAdsGenAI/) (2024-01-23): Bridging production ad-ranking personalization with multimodal generative synthesis using Gemini Pro Vision.
- [Exploring Deepfakes - Deploy Powerful AI Techniques for Face Replacement and More 🥸](https://jigyasa-grover.github.io/ExploringDeepfakes/) (2023-06-20): Technical breakdown of generative synthetic media, deepfake synthesis architectures, detection limits, and organizational provenance defenses.
- [Top 5 trends in Artificial Intelligence for 2023 📈](https://jigyasa-grover.github.io/top-5-trends-in-artificial-intelligence-for-2023/) (2023-02-13): Reflections with SaugaTalks on five defining shifts shaping enterprise AI adoption: generative models, data governance, explainability, spatial computing, and low-code ML.
- [My top 3 papers in ML for 2022 👀](https://jigyasa-grover.github.io/Top3MLPapers/) (2022-08-13): Practitioner notes on foundational machine learning papers shaping modern representation learning and production recommendation systems.
- [Data Correction with Cleanlab 🧼](https://jigyasa-grover.github.io/DataCleaning-Cleanlab/) (2022-08-13): Automatically detecting and fixing mislabeled training examples in real-world datasets using confident learning and Cleanlab.
- [Human Data Annotators - Lifeline of the AI pipeline!](https://jigyasa-grover.github.io/HumanDataAnnotators/) (2022-08-03): Designing fair, high-signal human annotation rubrics, inter-annotator agreement workflows, and quality feedback loops for ML datasets.
- [Differential Privacy for Privacy-Preserving Machine Learning 🔐](https://jigyasa-grover.github.io/DifferentialPrivacy/) (2022-07-30): Mathematical intuition and engineering trade-offs behind epsilon-differential privacy, calibrated noise injection, and privacy budgets in ML.
- [Privacy Preserving Machine Learning 🔐](https://jigyasa-grover.github.io/PrivacyPreservingML1/) (2022-07-10): Part 1 of a technical series examining membership, property, and model-inversion attacks alongside modern privacy-preserving ML defenses.
- [Data Lakes for Big Data ⛵️](https://jigyasa-grover.github.io/DataLakes/) (2022-01-25): How data lakes, schema governance, and feature lineage underpin reliable downstream machine learning pipelines.
- [Demystifying Data Quality in Machine Learning 🤓](https://jigyasa-grover.github.io/Data-Quality/) (2021-09-09): Quantifying completeness, consistency, timeliness, and validity across training and serving data pipelines.
- [Taking the Data-Centric route to Machine Learning 🚴🏻‍♀️](https://jigyasa-grover.github.io/Data-Centric-ML/) (2021-08-11): Why systematic dataset engineering outperforms endless hyperparameter tuning when taking ML models from benchmark notebooks to production.
- [Data Cascades in Machine Learning ⛲️](https://jigyasa-grover.github.io/Data-Cascades/) (2021-07-28): Understanding compounding upstream data issues—data cascades—in high-stakes machine learning deployments and how to prevent them early.
- [5 tips for polishing your Machine Learning dataset 🧼](https://jigyasa-grover.github.io/Polishing-Your-Dataset/) (2021-07-18): Practical preprocessing, outlier handling, and signal normalization techniques from Sculpting Data for ML.
- [BOOK LAUNCH 🎉</br>Sculpting Data For ML - The first act of Machine Learning](https://jigyasa-grover.github.io/BOOK-LAUNCH-Sculpting-Data-for-ML/) (2021-01-24): Introducing Sculpting Data for ML: The First Act of Machine Learning—why we wrote a hands-on guide to dataset curation and data-centric engineering.
- [Research Reading: Approaches to click-through rate estimation (3/x) 🤓](https://jigyasa-grover.github.io/research-reading-approaches-to-click-through-rate-estimation-3-x/) (2019-12-24): Part 3 of the CTR estimation research series: dissecting Alibaba's AAAI 2019 Deep Interest Evolution Network (DIEN) for modeling temporal user interest shifts.
- [Research Reading: Approaches to click-through rate estimation (2/x) 🤓](https://jigyasa-grover.github.io/research-reading-approaches-to-click-through-rate-estimation-2-x/) (2019-12-24): Part 2 of the CTR estimation research series: examining Yahoo! Research's Deeply Supervised Semantic Model for sponsored search click prediction.
- [Research Reading: Approaches to click-through rate estimation (1/x) 🤓](https://jigyasa-grover.github.io/research-reading-approaches-to-click-through-rate-estimation-1-x/) (2019-12-24): Part 1 of the CTR estimation research series: analyzing Microsoft Bing's production model-ensemble architectures for large-scale search ad click prediction.
- [Perceiving Python Programming Paradigms 🐍](https://jigyasa-grover.github.io/perceiving-python-programming-paradigms/) (2019-10-20): In-depth guide to imperative, procedural, object-oriented, and functional programming paradigms in Python—featured among Opensource.com's Top Articles.
- [Dive into Open Source – Go beyond coding 👩🏻‍💻](https://jigyasa-grover.github.io/dive-into-open-source-go-beyond-coding/) (2018-03-30): How engineers, technical writers, designers, and researchers create high-leverage impact in open-source ecosystems beyond writing code.
- [Keynotin’ at Open Source 101 🎙](https://jigyasa-grover.github.io/keynotin-at-open-source-101/) (2018-02-27): Notes and takeaways from delivering the keynote at Open Source 101 on sustaining open-source communities and lowering contributor barriers.
- [Paris, France to Maribor, Slovenia .. and beyond – Chronicles of CORMAS Pharo 🤠](https://jigyasa-grover.github.io/paris-france-to-maribor-slovenia-and-beyond-chronicles-of-cormas-pharo/) (2017-09-08): Engineering chronicle on modernizing the CORMAS multi-agent ecological simulation platform in Pharo Smalltalk across Paris and ESUG Maribor.
- [Winning Red Hat Women in Open Source Academic Award 2017, The Summit, A week in Boston and more …](https://jigyasa-grover.github.io/winning-red-hat-women-in-open-source-academic-award-2017-the-summit-a-week-in-boston-and-more/) (2017-05-31): Receiving the 2017 Red Hat Women in Open Source Academic Award at Red Hat Summit in Boston and reflections on open-source mentorship.