Participating in the Building Agentic AI Application Challenge with Decoding Data Science pushed me to focus on building AI that solves a tangible, high-stakes problem: delivering trustworthy agronomic guidance to farmers. That effort resulted in AgroVn AI—an autonomous field assistant designed around domain safety and reliability: ◄ Agentic Core: Built with LangGraph and Qwen-27B on Groq for fast, low-temperature reasoning. ◄ Grounded Knowledge: Combines a 13-crop verified calendar with extension search and weather advisory fallbacks. ◄ Zero-Hallucination Guardrails: Prioritizes Integrated Pest Management (IPM) and avoids fabricating chemical dosages. Huge thanks to coach/mentor- Mohammad Arshad sir for organizing such an execution-focused challenge. 👉 Live demo on Hugging Face: https://lnkd.in/gPur2gRx 🔗 Link of the official post : https://lnkd.in/p/eXWd6Jy5 DDS Business Circle #AICommunity #AgenticAI #AgTech #BuildInPublic #AIBuilders #GenAI #LangGraph #Groq #HuggingFace #LLMs #Python #AIChallenge #AIagents

I’m really happy to share that Figuro made it into the Top 21 Finalists in the DDS Building an Agentic AI Application Challenge! I’m grateful for the opportunity and excited to keep improving Figuro and learning throughout the challenge. Thank you to everyone in the DDS community for the support!

Introducing Nevin DDS AskHR AI, my enterprise HR chatbot built as part of the AI Accelerator Bootcamp @decordingdatascience and Mohammad Arshad Employees can use the chatbot to get answers to questions regarding hours worked, annual leave balances, requesting leave, working remotely, and raising workplace issues. Powered by LlamaIndex, RAG, OpenAI Embeddings, GPT4o mini, and Gradio, I was able to link its answers to human resources policy documentation. It was rewarding to see the chatbot respond with answers to actual HR questions after troubleshooting document retrieval, validating answers with company policy, and debugging deployment. Thanks to DDS for allowing me to apply what I learned into a functioning app. I plan to continue to iterate on it by surfacing source citations and creating more channels for user feedback. 🔗 The DDS HR Chatbot Link: https://ddshrent-j11q.onrender.com/ 🔗 The DDS HR Chatbot Link: DDS AskHR AI - a Hugging Face Space by Nevin-Pinto #DecodingDataScience #DDS #AIAcceleratorBootcamp #GenerativeAI #RAG #LlamaIndex #OpenAI #HRTech #AIProjects

As part of the DDS AI Accelerator Bootcamp, I developed and deployed DDS AskHR AI, an enterprise HR chatbot designed to provide employees with clear and reliable answers grounded in approved HR policy documents. The solution uses Retrieval-Augmented Generation (RAG) with LlamaIndex, OpenAI embeddings and an OpenAI LLM, combined with a Gradio interface and deployment on Hugging Face Spaces. (Hugging Face) Key capabilities include retrieving information across employee, leave, FAQ, and remote-work policies; identifying relevant policy sources; maintaining conversational context; and applying guardrails to avoid unsupported or sensitive answers. I also tested the chatbot across multiple HR scenarios, including working hours, annual leave, remote work, and out-of-scope salary questions, refining both retrieval accuracy and responsible AI behavior. This project gave me valuable hands-on experience in taking an AI solution from problem definition → RAG development → testing → refinement → deployment. A great step forward in strengthening my practical skills in AI, RAG, enterprise automation, and responsible AI development. #ArtificialIntelligence #GenerativeAI #RAG #LlamaIndex #OpenAI #HuggingFace #Gradio #AIEngineering #ResponsibleAI #HRTech #EnterpriseAI #Automation #AIBootcamp #DecodingDataScience #DDS #AIBuilders

I’m thrilled to announce that I’m participating in Decoding Data Science’s (DDS) AI Accelerator Bootcamp. Over the course of these 3 days, I’ll be attending the live sessions where we’ll learn how to bring an AI idea to life through a working prototype. My project idea is to build a DDS Enterprise HR Chatbot. To get started, I defined the HR problem, who the users are, where it will get knowledge from, what the chatbot will be able to do, what it can’t do, some sample questions, and how we’ll know if we’ve been successful. During the accelerator bootcamp, we’ll be using the CPE Framework to learn how to go from idea to valuable AI tool. 💡 Concept - Define the problem, your users, and your goals. 🛠️ Prototype - Build the idea into a functioning AI solution. 📊 Evidence - Test your results and evaluate for improvements. Building effective AI solutions starts with being clear on the problem you’re trying to solve. Thanks to @DecodingDataScience, Mohammad Arshad and the DDS community for another opportunity to learn by doing! #aiaccelerator #ArtificialIntelligence #AIBuilder #EnterpriseAI #HRChatbot #AIAgents #generativeAI #rag #AIEngineering #BuildInPublic #LearningByBuilding #DecodingDataScience