Jayanta Meslova

Currently developing AI-powered Question Generation System leveraging GPT-4 Omni and OpenCV for pattern recognition, with AWS S3 integration for scalable data management; implementing AI agents using Swarm framework for contextual question generation; hosting image and video models on RunPod GPUs; building a fashion-visualization platform synthesizing user-uploaded clothing items with AI-generated imagery; constructing AI-powered video generation workflows using GPT-4o, ElevenLabs for AI voice, Stable Diffusion for image creation, and WGAN-2 for animation.

Developed an AI Financial Advisor Platform by integrating LangChain, LLM Agents with Crew AI and AutoGen; implemented embedding with OpenAI ada-002 and Pinecone DB for vector storage; used IBM Watson API for call transcription and Twilio REST API for advisor-user communications; built a loan prediction model using NumPy and Pandas with SGD-based training; designed a specialized recommendation system to onboard new and existing customers via a two-model ATS optimization.