Google for Startups

The best analysts don’t just answer questions. They surface what matters before anyone asks. Gravity built Orion on Gemini Enterprise Agent Platform to do exactly that. ⚡️💻 Orion connects data across BigQuery, Looker, and Google Workspace, continuously monitors for meaningful changes, and recommends next steps directly inside the Gemini app. Instead of searching for insights, teams get them proactively. Start building your AI analyst today:
https://goo.gle/4v9vu9u

Big news out of #GoogleForBrazil! We are officially expanding the Google AI Futures Fund to Brazil, partnering with venture capital leader MONASHEES to launch the Gama Fund. Through this partnership, the Google AI Futures Fund is co-investing in and co-developing an inaugural cohort of elite deep tech startups, giving Brazilian founders a launchpad to build globally competitive companies. What the Gama Fund brings to selected startups: 💰 Up to $2M in recommended total co-investment per startup. ☁️ $350,000 in Google Cloud and Gemini credits. 🧠 Early access to advanced Google DeepMind models (like Gemini, Imagen, and Veo). 🛠️ Direct co-development with Google engineering teams and dedicated startup managers. 🏢 Physical access and support at the new Google Campus hub at IPT Open. This collaboration gives founders the tech and capital to scale, while their feedback helps us build better, more localized AI solutions for the region and the world. 👉 Ready to build the future of deep tech? Apply now: https://goo.gle/4xomMFH

Looking forward to being back in London next week and participating in the Builders Breakfast with Lightspeed and Google DeepMind on Wednesday, June 10th. If you're building a startup in Europe, you don't want to miss this! I'll be chatting with Clément Farabet (VP, AI Engineering at Google DeepMind) and Antoine Moyroud (Partner, Lightspeed), where we'll discuss what's actually happening at the frontier right now, the infrastructure we're building, the models we're scaling, and what it means for the people building on top of them. If you're deep in the weeds on something ambitious, and want a small room of people who get it then register here: https://lnkd.in/egm2ci8W Darren Mowry, Nikao Yang, Michael Sandwick, Erin Schoch, Oksana Stowe

Como a sua startup usa a IA para escalar os negócios? 💡 No Web Summit Rio, Milena Leal (Google Cloud) e Graciela Kumruian (CEO da Netshoes) deixaram um aviso: inovação sem execução é só projeto de gaveta. Aqui, Milena Leal, Country Manager Brazil, no Google Cloud, lembra que o nosso país tem talento e infraestrutura para liderar o jogo global. 📈 Arrasta para o lado e fique por dentro dos destaques da conversa que aconteceu esta semana. E conta nos comentários: qual a realidade da IA na sua startup? 👇 #GoogleCloud #Netshoes #WebSummitRio #Liderança #Inovação #IA #FatorHumano

Build AI agents that run directly on your users’ devices. No cloud required. 📱⚡️ Gemma 4 makes on-device AI agents practical to ship. No constant server calls, lower operating costs, and sensitive user data stays on-device. This opens up new product experiences for teams building around privacy, real-time responsiveness, or reliable offline functionality. Swipe through to see what you or your dev lead can build with Gemma 4 today. 👉

Customer feedback only creates value when it changes what you build. With Gemini, you can turn thousands of survey responses into structured product insights in minutes. 👉 ⚡ Collect feedback through Google Forms 🧠 Have Gemini automatically categorize responses by topic and sentiment 📊 Review organized insights directly in Google Sheets No manual tagging. No digging through spreadsheets. Just a clear view of what customers love, where they're getting stuck, and what deserves your team's attention next. How are you turning customer feedback into product decisions? 👇

De 8 a 11 de junho, o ecossistema de startups estará reunido no Web Summit Rio 2026 para debater o futuro dos negócios. ✨ E é claro que o Google for Startups também estará lá! Anote na agenda dois papos imperdíveis para o seu negócio (que também serão transmitidas online no YouTube do evento): ⏱️ 09/06: Milena Leal (Google Cloud) e Graciela Kumruian (Netshoes) mostram como dar saltos práticos no negócio com IA. ⏱️ 11/06: Fábio Coelho (Google) e Mate Pencz (Loft) debatem estratégias reais de captação de recursos e sustentabilidade. E confira as masterclasses imperdíveis que preparamos para quem se inscrever no Web Summit 2026. Ah, e não deixe de conhecer nosso estande interativo. 👉 Acesse aqui programação completa: https://goo.gle/4vfVK21 Acompanhe o bate-papo do dia 09/06 aqui: https://goo.gle/4afbedW acompanhe o bate papo do dia 11/06 aqui: https://goo.gle/4vVouwX #GoogleForStartups #GoogleCloud #WebSummitRio #WebSummit2026 #StartupsBrasil #Inovação

700+ founders, developers, partners, and Googlers from Google for Startups and Google DeepMind came together at #GoogleIO After Hours to share what they're building with Gemini and Gemma. 🚀 A few themes stood out: ⚙️ Gemini 3.5 is helping teams build more capable AI workflows with stronger tool calling 🧠 Gemini is powering highly personalized user experiences 🛠️ Gemma's open models are giving developers more flexibility to build and deploy AI applications 📱 Better tool support is making AI-powered experiences feel seamless across products and platforms Which use case are you most excited about?

High latency in your AI agent usually isn’t a model problem, it’s a retrieval bottleneck. If you’re reprocessing the same documentation on every query, you’re adding delay and cost to every interaction. ⚠️ Enable context caching on the Gemini Enterprise Agent Platform so your model doesn’t have to reprocess the same data every time. You go from seconds-long delays to near-instant responses, with lower inference cost. ⚡️✅ Where are you still seeing delays in your agent? 👇

Your AI is answering questions it can’t verify. Here is the blueprint for a trusted AI agent using Gemini Enterprise Agent Platform (previously Vertex AI) 👉 🧠 Anchor to your data: Use RAG and Agent Search so responses come from your docs 🔍 Ground in real time: Check outputs against live Google Search before users see them If you don’t control the source, you don’t control the answer. Where in your product are responses still going unsourced?