
1
1
Bangladesh
3+ years of experience
I’m Esha, a passionate developer who loves building real-world solutions with code. My interests span Python, C/C++, Java, MySQL, HTML, CSS, and emerging areas of AI and embedded systems. I enjoy taking on challenging problems—whether it’s designing algorithms, crafting full-stack projects, or experimenting with microcontrollers and hardware. I’m currently exploring machine learning, data-driven applications, and digital systems design to deepen both my theoretical and hands-on knowledge. I love learning, improving, and creating things that actually help people. Outside academics, I also enjoy participating in hackathons and tech competitions where I get to test my skills and grow through experience. I’m always excited to connect, collaborate, and work on meaningful projects.

AI Reality Integrity Engine is a fully autonomous fact-checking and truth-verification system designed to combat misinformation in real time. The project combines advanced natural language understanding, claim extraction, evidence retrieval, and NLI-based reasoning to evaluate the accuracy of any statement a user submits. The system works in five stages: Claim Extraction — It identifies factual statements inside user-provided text. Evidence Retrieval — It optionally searches the web for reputable sources using APIs like Tavily. AI Verification — Using state-of-the-art Groq models (Llama 3.3 70B / 8B), it performs deep reasoning to classify each claim as Supported, Contradicted, or Unverified. Verdict Scoring — It generates a truth score, confidence level, and structured explanation. Summary Generation — It outputs a concise overview for the end user. The project includes a FastAPI backend, a responsive Streamlit frontend, a fully cloud-deployed architecture, and modular design suitable for integration with news platforms, academic tools, and safety applications. Its mission is to make truth verification fast, accessible, and reliable for everyone.
19 Nov 2025