Still processing this one honestly. I built Vault in 8 days, a personal finance tracker with Frank, a blunt AI raccoon who reads your actual spending numbers and tells you exactly what's wrong with your financial decisions. No generic advice. No sugarcoating. Just your data and a raccoon with strong opinions. What started as a blank Python file turned into a fully deployed app with real auth, a real database, real charts, and an AI persona that genuinely made people laugh during testing. That alone felt like a win before the results even came out. Huge thank you to the Decoding Data Science team and the judges for putting this together. This kind of challenge teaches you more in 8 days than months of tutorials ever could, and I mean that genuinely, not just as something nice to say. Congrats to everyone who built and submitted. The standard was high and it pushed me to do better. 🚀 🔗 Live app: https://vault-finance.streamlit.app/ 📁 GitHub: https://github.com/bluejay-19/Vault #AIChallenge #BuildInPublic #Vault #DecodingDataScience #DDSBusinessCircle #BuildAIAppChallenge #Python #Streamlit #Groq

Today, Planit is a fully deployed AI-powered web app that takes your goal, your deadline, and your experience level and turns it into a clear, personalized day-by-day action plan in seconds.✨ What Planit can do: 🚀 Generate a personalized day-by-day action plan for any goal ⏱️ Respects your daily time limit — no overwhelming schedules 📚 Free resources included every single day 🎯 Smart task ordering — always introduce before you practice 🌙 Handles edge cases — vague goals, short deadlines, small goals 🧘‍♀️ Fitness-specific planning with rest days and realistic sessions ⭐ Optional bonus days when your goal is smaller than your deadline ✨ Personalized intro and outro — never generic, always motivating Here's how the 8 days went: Day 1: Set up Python, VS Code, installed libraries and pushed first commit Day 2: Built core API integration and basic Streamlit UI Day 3: Implemented full structured prompt and drafted 20 test cases Day 4: Tested edge cases and refined prompt output format Day 5: Built full UI with expanders, progress bar and styling Day 6: Added two-layer input validation, secrets handling and fixed bugs Day 7: Optimised prompt parameters and fine-tuned LLM interactions Day 8: Recorded demo, deployed live and submitted These 8 days have been an incredible learning experience. From prompt engineering to deployment, every day taught me something new. I’m looking forward to what I’ll build next. 💫 Planit: planit-ai.streamlit.ap🪐

Launch complete. Planit 🪐 is now live. Today was deployment day — taking everything built over the past week and putting it out into the world for anyone to use. Here’s what I got done: deployed Planit on Streamlit Community Cloud, set up the API key securely using Streamlit’s secrets manager, ran smoke tests on the live version, and checked performance and latency post deployment. Day 8 marks the finish line. Demo video creation and final submission prep underway ✨

Every year, millions of tourists visit Dubai, yet many still plan their trips using outdated blogs, dozens of browser tabs, and guesswork. So I built NovaDXB—an AI Concierge that thinks like a local Dubai expert. Instead of giving generic answers, it creates a personalized travel plan based on your budget, travel style, group, and trip duration. ✨ In seconds, it generates: 🗺️ A day-by-day itinerary 📍 An interactive map with live location pins 🍽️ Restaurant recommendations 💰 Real AED budget estimates 💡 Local insider tips Behind the scenes, NovaDXB uses an Agentic RAG architecture powered by LangChain, LangGraph, LlamaIndex, Pinecone, and GPT-4o-mini, with built-in security features like prompt injection detection, response caching, and rate limiting. 📊 Results: ✅ 20/20 evaluation score ✅ Tested across 3 LLMs ✅ Average cost: $0.000375 per query This project was built for the Decoding Data Science App Building Challenge, and it taught me how much goes into building a production-ready AI application—not just prompting an LLM. 🚀 Live Demo: https://huggingface.co/spaces/nipunkavindaAI/NovaDXB 💻 GitHub: https://github.com/NipunKavinda95/NovaDXB I'd love your feedback! If you were visiting Dubai, what feature would you want an AI travel assistant to have? #NovaDXB #AIAgents #GenerativeAI #RAG #LangChain #BuildInPublic

8 days ago I had a blank Python file and an idea. Today I have a deployed, working AI application that real people can open in a browser right now. ✨ 🥗 Cookable is an AI powered recipe assistant built around a problem most of us have faced. You open your fridge, you have ingredients, you have no idea what to make, and you end up ordering food anyway. 📝 You type in whatever is in your fridge, set your preferences, and the app returns 5 cookbook quality recipes with exact heat levels, timings, and visual cues. Not summaries. Actual instructions you can cook from. Here is what Cookable can do: 🍳 Generate 5 detailed recipes from whatever ingredients you have and pantry staples 🥙 Filter by cuisine, time available, meal type, serving size, and dietary restrictions 🛒 Get 10 smart ingredient suggestions with reasons to buy them 🧈 Generate 2 bonus recipes when you pick up new ingredients 🕘 One click history pills to reload any past search instantly 🌱 Clean earthy green and cream UI built entirely with custom CSS in Streamlit Here is what the 8 days actually looked like: 🫑 Day 1: Ideation. The idea came from standing in front of my fridge at midnight with eggs, a carrot, and half a block of cheese, and ordering food anyway. 🥦 Day 2: Environment setup. Groq API connected, Streamlit running locally, first test prompt returning a response. 🥒 Day 3: The hardest day. Writing prompts that return consistent structured JSON with cookbook quality steps every single time took far more iteration than expected. Small wording changes had a bigger impact on output quality than anything else. 🫛 Day 4: Evaluation and hardening. 20 test cases, three failure modes found and fixed, rate limit handling added, input validation tightened. 🍏 Day 5: Full UI. Custom CSS design system built entirely in Streamlit. Green and cream palette, hero image, history pills, bonus recipe section, sticky footer. 🥬 Day 6: Security. System and user message split, prompt injection guard, retry logic, JSON fallback parser, README cleanup.