CloudRun-AI-Suite

🌟 Inspiration We wanted to explore how far AI can go in automating the full developer workflow — not just coding, but also building, deploying, and scaling applications in the cloud. With Google Cloud Run and AI Studio, we saw an opportunity to bridge the gap between idea and production. CloudRun AI Suite was born to show that a single prompt can spin up a real, working, serverless application — in minutes, not days. 🤖 What it does CloudRun AI Suite converts natural language prompts into deployable, scalable serverless apps using Google Cloud Run and Gemini. It features three headline demos: Prompt-to-API Generator: Converts plain English API descriptions into REST services with OpenAPI specs and Dockerfiles. AI Resume Analyzer: Extracts insights from resumes (PDF/TXT), anonymizes data, summarizes candidate strengths, and scores profiles with Gemini. EduBot: Creates short educational modules and quizzes using Gemini, with deterministic output and an elegant web UI. All three are orchestrated through an AI-driven “App Canvas” — a unified interface to generate, deploy, and interact with each module seamlessly. 🛠 How we built it AI Studio (Gemini): Used for prompt-to-code generation and text understanding. Cloud Run: Deployed each service as an independent, autoscaling container. Firestore + Pub/Sub: Managed agent communication, app logs, and regeneration triggers. NVIDIA L4 GPUs: Accelerated inference workloads for model-backed APIs. CI/CD: GitHub Actions automates build → deploy cycles. Frontend: Built with React + TailwindCSS for a modern, responsive interface. Terraform: Automated Cloud Run, Firestore, and Pub/Sub provisioning. We used AI Studio to generate boilerplate code, then refined and connected services to work harmoniously across the cloud ecosystem. ⚡ Challenges we ran into Designing prompts that generated production-ready code without manual intervention. Integrating AI-generated APIs with real Cloud Run deployments under time constraints. Managing ephemeral storage and secure file uploads in the Resume Analyzer. Ensuring deterministic Gemini outputs for EduBot’s quiz generator. Synchronizing multiple services via Pub/Sub and CI/CD while maintaining clean infrastructure. 🏆 Accomplishments that we're proud of Built three complete AI-powered microservices deployed live on Cloud Run. Achieved a full “prompt-to-production” workflow with minimal human coding. Integrated GPU workloads for inference acceleration. Created a sleek, animated App Canvas UI that makes AI-assisted app development intuitive and fun. Demonstrated the power of serverless AI orchestration in a real-world hackathon setting. 📚 What we learned How to design effective AI Studio prompts that translate natural language into structured, working applications. The value of modular architectures for multi-agent AI workflows. How Cloud Run simplifies scaling, load balancing, and deployment automation. That serverless + AI truly democratizes software creation — anyone can become a builder. 🚀 What’s next for CloudRun AI Suite Add multi-agent orchestration with Google’s Agent Development Kit (ADK). Integrate Veo for video content generation in EduBot. Enable real-time code regeneration triggered by user feedback through Pub/Sub. Expand GPU-backed inference for advanced models (Gemma, Imagen). Launch a public “Prompt-to-Deploy” web platform so developers can create and share AI-generated microservices instantly.

Built With

• all