I’m building ProofCheck — an AI Evidence Reconciliation Agent. The idea is simple: Don’t just summarize documents. Prove whether they agree. The project combines: • RAG and structure-aware document retrieval • Agentic investigation and tool use • Deterministic validation for calculations and evidence conflicts • Structured evidence and provenance • Safety boundaries and human-in-the-loop review • Evaluation with golden test scenarios One principle has guided the architecture: The LLM decides what to investigate. Deterministic systems establish what the numbers actually say. Today I completed the first working UI connected to the real backend, including investigation progress, error handling, and reset/clear behavior. There is still work to do — especially around reliability and edge cases — but that is exactly what makes this challenge valuable: building, testing, finding weaknesses, and improving the system one step at a time. On to the next checkpoint. 🚀 #AI #ArtificialIntelligence #AgenticAI #RAG #LLM #Python #AIEngineering #BuildInPublic #DecodingDataScience #AIChallenge

ProofCheck is now moving from RAG into an agentic evidence-reconciliation system. Today I built: - 🔎 A bounded LLM investigation agent - 🛠️ Typed tools for evidence search and validation - 🧮 Deterministic reconciliation for quantities and calculations - 📋 Structured outcomes: SUPPORTED / REVIEW_REQUIRED / BLOCKED - 🛡️ Safety boundaries and controlled API retries The core engineering principle remains: The LLM decides what to investigate. Deterministic systems establish what the numbers actually say. Next: Day 5 — building the interface and connecting the full workflow. #AIEngineering #LLM #RAG #AIAgents #Python #BuildInPublic #DataScience#DecodingDataScience

They say the boring days are where real products get built. Day 2 of ProDiag AI was exactly that. 🛠️ Environment configured. Core systems coming online. Nothing flashy to show yet — but the engine is starting to turn. Every AI agent that looks effortless in a demo had a day like this behind it. 6 days left in the AI Application Building Challenge. Let's keep building. 🔥 Decoding Data Science DDS Business Circle Mohammad Arshad

Today I focused mostly on testing and fixing what I already built. I tested the AI Tutor and fixed the connection, tested Practice and fixed the repeated-question issue, tested Solve Together to make sure it actually guides me step by step, and checked Progress to see whether Figuro can detect my weak points from my learning activity. I also added the Quiz navigation to the app. Day 2 was basically: test → find problems → fix them → test again. #Figuro #BuildInPublic #AI

Excited to share the first deployed prototype of Nevin Chatbot – an AI Math Learning Assistant. 🥳😎 The mission with this tool is provide students the ability to understand the math being done, not just the answer! We do this with step-by-step learning, explanations of concepts and math, and most importantly by focusing on the student experience. Prototype UI built ✔️ Github repo set up ✔️ Deployed website via Github pages ✔️ ⏰ Next Up - Build AI brain and hook up the LLM model 🧠 Live Demo Website Link: https://lnkd.in/d5N4BSth Live Project Repository Link: https://lnkd.in/d8EUsUPM Thanks to Decoding Data Science, Mohammad Arshad and the rest of the DDS community for pushing us to learn by creating! 🙌🏽 hashtag#DDSNextGen hashtag#DDSAmbassador hashtag#DecodingDataScience hashtag#ArtificialIntelligence hashtag#StudentBuilder hashtag#LearningByBuilding