The Ambassador Role
Google selects campus ambassadors to drive AI adoption at universities — students who can translate cutting-edge API capabilities into workshops that other students can actually follow. The role has three components: organize events, create educational content, and actively support student projects that use the Gemini API.
At GTU, this means something specific. Gujarat Technological University is a large state university with strong engineering enrollment and real hunger for practical skills. But most students haven't used a production AI API. They know ChatGPT as a chat interface; they don't know how to call an endpoint, handle a response, or wire a model into an application they're building. That gap is exactly what the ambassador program exists to close.
What I Do
The role combines facilitation, mentoring, content creation, and community infrastructure. None of these in isolation — all four together, running in parallel across the semester.
Workshop Facilitation
Design and run hands-on Gemini API sessions — from first API key to a working prototype. Structured for students who have never made an API call before. Every session ends with something built, not just a slide deck absorbed.
Project Mentoring
Help students integrate Gemini into hackathon projects, final year projects, and side builds. This is 1:1 and small-group work — debugging API calls, explaining token limits, suggesting prompting strategies that actually work.
Content Creation
Tutorials, annotated code examples, and step-by-step guides written in plain language. The goal: a student who missed the workshop can still follow along and get to a working result. Reusable, shareable, versioned on GitHub.
Community Building
Discord server for GTU AI builders, peer-to-peer study groups, and async support channels. The community persists between workshops — students share what they've built, ask questions, and pull each other forward.
Workshops & Events
The workshop series is progressive — each session builds on the last, so students who attend regularly move from zero to building real features with Gemini. Drop-in sessions are also welcome; each workshop is designed to stand alone.
Gemini API Basics
First contact with the API. Getting a key from Google AI Studio, making the first call, reading the response, understanding the model parameters. By the end: every student has a working Python or Node.js script that calls Gemini and does something useful.
Building with AI
Practical project sessions — taking the API call from session 1 and wiring it into a real application. Text summarization tools, document Q&A, content generators. Focus on the integration layer: how to call an API inside a web app, handle errors gracefully, and build something a real user could use.
Hackathon Prep
Pre-competition workshops timed to upcoming hackathons. Speed-building sessions: scoping an AI feature in 2 hours, prompt engineering for specific output formats, working with multimodal inputs. The goal is hackathon-ready skills, not polished code.
Open Office Hours
Weekly async drop-in. Students bring their own projects, stuck points, and questions. No agenda — just help with whatever they're building. This is where the real learning happens: messy, context-specific, actual problems.
Projects I've Helped Build
The measure of an ambassador program isn't attendee count — it's projects shipped. These are real things students and I have built using Gemini, ranging from hackathon prototypes to tools now in production.
Permit Objection Generator (Open Permit)
Started as AFOG — a 10-hour hackathon build using Gemini to generate legally-structured objection letters from planning permit documents. Won Code4Compassion. Now rebranded as Open Permit, live in 8+ countries. The original Gemini integration is still the core of the letter generation pipeline.
LEWS Early Warning System
Built for the EA Animal Welfare Hackathon 2025. A multi-variable scoring system for assessing lock-in risk in emerging animal farming technologies. Gemini used for natural language interpretation of policy documents and research papers feeding the scoring pipeline.
Student Project Chatbots
Multiple students have shipped chatbots for final-year projects using Gemini — customer support bots, college FAQ assistants, department-specific query handlers. The workshop-to-project pipeline is working: students attend, learn the API, and build within 2-3 weeks.
Resume Parsing Tools
Several students built resume-to-structured-data pipelines using Gemini's document understanding. Used for college placement cells and student portfolio generation. Gemini's ability to handle unstructured PDF content made this a practical first project for new API users.
Why This Matters
"Most GTU students have never made an API call. Lowering that barrier — even by one workshop — multiplies the number of people who can build with AI."
There is a real access gap in AI education at Indian universities. The tools exist. The APIs are free or cheap. The documentation is available in English. But none of that helps a student who doesn't know where to start, doesn't have a senior to ask, and has never seen anyone make a live API call in front of them.
That's what the ambassador role provides: a trusted peer who has used the tools in real projects, can answer questions without gatekeeping, and runs workshops in a format that's actually accessible. The technical bar at these workshops is deliberately low — the goal is the first successful API call, not architectural mastery.
Once a student has made one API call, the trajectory changes. They can read the docs. They can debug. They can build. The workshop isn't the destination — it's the unlock.
Key Lesson
Teaching is learning twice. Running Gemini workshops forced me to understand the API deeply enough to explain it simply — which means understanding it at a level that's actually useful for building production features, not just demos. The ambassador role has made me a better developer because it requires genuine comprehension, not surface-level familiarity.