For builders

Gemini and Gemma

What I love and hate about building with Gemini

I have used Gemini and Gemma across most of my hackathon projects. These are the experiences that stuck with me.

36 projects 34 competitions Gemini API, Gemini Live native audio, structured output, Gemma

What I loved

  1. Native audio was fast enough that interrupting it felt natural

    Room Sense at Gemini 3 Hackathon, Jan 2026

    What I was trying to do
    Room Sense is a rehearsal room for stand-up comedians. You upload a set, then talk it through out loud with four coaching voices, one each for premise, structure, cutting words and delivery. Comedy runs on timing, so the voices had to answer at the speed of a conversation, not a chatbot.
    What happened
    Instead of the usual chain of transcribing speech, sending text to a model and turning the reply back into audio, Room Sense streams raw microphone audio from the browser straight into Gemini's native-audio model over a live connection, and gets audio back. The replies came back fast enough that I could interrupt it and the exchange still felt like a conversation. The same model also picked up on hesitation and a lack of conviction in a performer's delivery in a way a text model never could.
    Why I loved it
    The transcribe-then-generate-then-speak pipeline is too slow for this and kills the joke. Without that speed there is no jam session, only a delay. Native audio is what made the live half of Room Sense possible at all.

    From the Room Sense project story

  2. Gemini 3 could tell the premise a joke states from the one it actually rests on

    Room Sense at Gemini 3 Hackathon, Jan 2026

    What I was trying to do
    Before the live session, Room Sense sends the written set to Gemini 3 Flash to build a map of the material: what each bit claims to be about, whether the setup and punchline hold together, and where words can go. That map is what the live coaching voices refer to afterwards.
    What happened
    Every joke has the premise it announces and an assumption underneath that the audience has to share for it to work. Gemini 3 separated the two. Its reading of what a bit was really about, as opposed to what it said it was about, was surprisingly human-like, the kind of note a veteran comic gives in a writers' room, and it read as genuine artistic insight rather than a summary.
    Why I loved it
    That is the note that is hardest to get from a machine and the most useful to get from a person. It is also why Room Sense pairs a slower reasoning model for the map with a fast one for the conversation: the map is where the insight lives.

    From the Room Sense project story

What I hated

  1. It fabricated citations routinely, even inside a valid JSON schema

    Signal Packet API at Brainwave 2026 – X402 Blockchain Track, Jul 2026

    What I was trying to do
    Signal Packet API reads a hackathon submission, the description, the README and the docs, and reports which claims those documents actually back up, quoting the exact passage as proof. Judges get about ten minutes per entry, and sorting "asserted" from "documented" is mechanical work a machine should do for them. I asked Gemini 3.5 Flash for structured output: a JSON answer in a fixed shape, one entry per claim, each with a verbatim excerpt from the source.
    What happened
    The JSON came back in the right shape every time, and the excerpts inside it were regularly invented. Quotes that appeared nowhere in the README or docs I had fetched, presented as evidence. Not occasionally; routinely.
    Why I hated it
    A schema guarantees the shape of an answer, not its truth. I ended up treating the model's output as a proposal rather than a finding: the server checks every cited excerpt word for word against the fetched source, drops the ones that fail, downgrades the claim to "unclear", and reports the corrections to the caller instead of hiding them. Builder Signal, Judge Ops and Signal Review Reel all needed the same check. This is not a claim that Gemini "hallucinates" in general; it is what happened every time I asked it to quote evidence.

    From the Signal Packet API project story

Everything I built with Gemini and Gemma

36 projects from the participation archive, newest first.

  1. Clarity

    Clarity finds the one vague claim worth asking about, then uses a short adaptive phone call to turn it into a clearer fact, backed by the candidate’s own words.

    CALL-E: Your Code Is Calling

  2. Audition Agent

    Audition Agent turns an AI-generated character into a human casting workflow, creating auditions, organizing applicants, and handling follow-up while keeping casting decisions human.

    Agents for Humans Hackathon

  3. Payoff

    Payoff helps creators design what an audience should feel, then uses real audience reactions to discover whether the story actually got them there.

    The WebMCP Challenge

  4. Builder Signal

    Turns a technical builder’s real experience into evidence-backed, editable applications tailored to hackathons, fellowships, accelerators, and other selective programs.

    Build with Gemini XPRIZE

  5. Judge Ops

    Helps judges, recruiters, and program reviewers evaluate submissions faster with structured, evidence-backed analysis while keeping scores and final decisions human.

    Build with Gemini XPRIZE

  6. Beatboard

    Beatboard turns a comedy idea into a shoot-ready storyboard with timed beats, generated art, voiced dialogue, direction, and dependency-aware revisions in minutes.

    Agentic Cinema: The Blockbuster Hackathon

  7. Signal Packet API

    Send in a project write-up and get back a report on which of its claims the README and docs actually back up, with quotes as proof. Costs 5 cents per check, paid automatically with x402.

    Brainwave 2026 – X402 Blockchain Track

  8. Signal Review Reel

    Give it a project description and its demo video. It watches the video and makes a short narrated clip showing which claims the demo actually proves, which it skips, and which it contradicts.

    Backblaze Generative Media Hackathon

  9. Signal Evidence Graph

    It maps a project's files to the claims they support in DataHub, then an AI agent reads that map and reports which claims the evidence proves. When it can't tell, it says so instead of guessing.

    Build with DataHub: The Agent Hackathon

  10. First Session

    First Session helps NYC parents find a Medicaid teen therapist by surfacing the highest-leverage contact path first. One email can reach 15 therapists. Fail fast, not by chasing the perfect match.

    World Product Day: Everyone Ships Now

  11. TradeCast

    Self-improving AI trader: Bull, Bear & Risk agents debate, paper-trade via Alpaca, auto-publish explainer Shorts to YouTube, and rewrite their own prompts by querying Arize Phoenix traces over MCP.

    Google Cloud Rapid Agent Hackathon

  12. AlreadyAsked

    Semantic duplicate-post detection for Reddit mods. Catches paraphrased reposts that keyword bots miss, using sentence embeddings plus an LLM verification step to keep false positives low.

    Reddit Mod Tools and Migrated Apps Hackathon

  13. Fan Meaning Map

    Discover what kind of Team USA fan you are. Gemini decodes 4 quick answers into your fan archetype, your athlete matches, and the surprising reasons others see sport differently.

    Team USA x Google Cloud Hackathon

  14. Memory Moment

    Helps caregivers find the right words to say to someone they love with memory loss, in the moment it matters.

    The Gemma 4 Good Hackathon

  15. First Impression

    See how the same face gets judged completely differently at first glance.

    Build with MeDo Hackathon

  16. Clinical Decision Room

    A virtual tumor board for everyday clinical decisions. Four reasoning frameworks analyze the same patient in parallel, surface where they disagree, and show exactly what would flip the decision.

    Agents Assemble - The Healthcare AI Endgame

  17. Sacrifice Equation

    A way to see how different perspectives judge your idea, and what tradeoffs make it succeed.

    ZerveHack

  18. Agent Identity Manager

    A control center for AI agent identity. Assign roles, enforce permission boundaries via Auth0 Token Vault, and make every credential exchange and access decision visible in real time.

    Authorized to Act: Auth0 for AI Agents

  19. Persona Fork

    Take one idea, cook it different ways, and see what people actually like.

    Zero to Agent

  20. Room Sense Live

    Room Sense Live lets comedians test jokes with a live AI audience that laughs, reacts, and shows how different crowds interpret the same material.

    Gemini Live Agent Challenge

  21. Drift Guard

    Drift Guard turns every merge request into a lightweight architecture review, detecting structural drift before it becomes technical debt.

    GitLab AI Hackathon

  22. Audience Mind Split

    An interactive audience simulator that shows how different personas interpret the same line and lets you talk back to them in real time.

    Gemini 3 NYC Hackathon

  23. Backlash Predictor

    A multimodal AI tool that scans campaign materials to simulate audience reactions, surface backlash risks before launch, and suggest targeted rewrites without flattening the creative voice.

    DeveloperWeek 2026 Hackathon

  24. Cook It

    Cook It identifies difficult ingredients or tools in the video and offers tiered substitutions based on the user's reality.

    RevenueCat Shipyard: Creator Contest

  25. Learn It

    Learn It turns coaching and tutorial videos into guidance that fits your time. Paste a link and get a deep dive, a quick overview, or the one thing that matters right now.

    RevenueCat Shipyard: Creator Contest

  26. Tiny Crossword

    A tiny 5×5 crossword generated by AI using daily or player-chosen themes. Quick to solve, easy to play, and fun to race against puzzles created by the community.

    The Reddit Daily Games Hackathon

  27. Play Console

    A physical console for orchestrating live AI improv scenes. Shape energy, pacing, and relationships in real time, and jump into the scene yourself without typing prompts.

    DevStudio 2026 by Logitech

  28. Snipper

    Transforming raw, long-form comedy footage into high-engagement social media content by leveraging the multimodal reasoning of Gemini 3 Flash.

    Build your Flutter Butler with Serverpod

  29. Laugh Spikes

    Laugh Spikes treats stand up comedy like sports game tape by aligning audience laughter to transcript lines, visualizing joke impact, and generating data grounded AI feedback to improve a set.

    Hex-a-thon

  30. Room Sense

    An AI native performance lab for comedians where you workshop your set in real time with four specialized voice agents powered by Gemini 3 and 2.5.

    Gemini 3 Hackathon

  31. Timing Coach

    AI coach that teaches natural speech timing in 15 languages through real-time feedback and rhythm visualization, so you sound human, not textbook, across borders.

    VisaVerse AI Hackathon

  32. SixView

    SixView helps lender teams analyze complex loan agreements from six perspectives, surfacing risks, flexibility, and negotiation hotspots in minutes instead of weeks.

    LMA EDGE Hackathon

  33. YesAnd

    AI improv partner you can talk to in real time. Practice scenes with natural voice interaction and get instant coaching powered by Gemini. Built with ElevenLabs Conversational AI + Google Cloud.

    AI Partner Catalyst

  34. Business Opportunity Finder

    Spot high potential businesses for any neighborhood by analyzing local competitors, customer pain points, and unmet demand using Yelp data enhanced with Gemini AI.

    Yelp AI API Hackathon

  35. Roastify

    An AI that roasts your photos and turns them into scroll-stopping, laugh-worthy captions in seconds.

    Cloud Run Hackathon

  36. Meme Roaster

    Compete with AI in a meme roast challenge. Create your funniest line, and let Reddit decide who delivers the better

    Reddit and Kiro: Community Games Challenge