I'm always passionate about exploring machine learning, creating interactive educational tools, and pushing the boundaries of what we can build with AI. 💡 Quick question for you all: What project or AI tool are you currently building or experimenting with this week? Let’s connect and share insights in the comments.

I used to think I had to choose between coding and no-code. After building personal projects in Visual Studio Code and combining them with AI coding tools like Cloud Code, I realized they complement each other. No-code helps me prototype quickly. Coding gives me full control and customization. AI speeds up both. The goal isn’t to pick one. The goal is knowing when to use each.

When I first started learning AI, I thought the most important thing was using the latest models and tools. But after working on real projects, I realized that people don’t really care which model you used. They care about whether you solved a real problem. AI is just a tool. Understanding people’s needs, listening to feedback, and improving your idea is what really makes a difference. That’s a mindset I’ll keep with me as I continue building. Build to solve, not just to impress.

This week taught me that building an AI product is not only about creating a good prompt or model. The real challenge is connecting everything together and making the experience reliable for users. While developing Forsa, I learned more about: • Building and improving an AI interview agent • Connecting the frontend, backend, and AI APIs • Supporting Arabic and English experiences • Testing voice, video, and text interview modes • Debugging authentication, deployment, and user-flow issues • Turning feedback into practical product improvements My biggest lesson this week: an AI idea becomes valuable only when it solves a real problem and works smoothly for the user. I’m excited to continue learning, building, and sharing my progress with the DDS community. #Week5 #DDSAmbassador #AgenticAI #ArtificialIntelligence #Forsa #BuildInPublic

Excited to share that I've successfully completed the AI Accelerator Bootcamp by Decoding Data Science, earning 10 CPD-approved learning hours. This wasn't just another AI course—it was a hands-on bootcamp that emphasized building real-world AI solutions from idea to deployment. During the bootcamp, I developed and deployed: ⚡ Lumen AI — Dubai Utility Bill Assistant Try: https://huggingface.co/spaces/nipunkavindaAI/LumenAI An AI-powered assistant that helps users understand DEWA electricity and water bills, including: ✅ Electricity & Water Slab Tariffs ✅ Fuel Surcharge Explanation ✅ Housing & Sewerage Fees ✅ VAT on Utility Bills ✅ Energy-saving Recommendations Built using LlamaIndex, OpenAI, Pinecone, and Hugging Face Spaces, enabling grounded, RAG-powered responses for utility-related queries. One of the biggest takeaways for me was strengthening a practical workflow for building production-ready AI applications—from knowledge base design and retrieval pipelines to deployment and user experience. Key skills strengthened 🔹 Retrieval-Augmented Generation (RAG) 🔹 LlamaIndex & LangChain workflows 🔹 Vector Databases (Pinecone) 🔹 Prompt Engineering 🔹 AI Application Deployment with Docker & Hugging Face Spaces 🔹 Building AI solutions for real-world business use cases A special thanks to Mohammad Arshad and the Decoding Data Science team for creating a learning environment that focuses on practical implementation rather than just theory. Receiving a CPD-certified completion certificate with 10 learning hours makes this achievement even more meaningful, reflecting continued professional development alongside hands-on AI engineering experience. Looking forward to building more AI applications that solve real-world problems.