Ashish Solanki

Background

About Ashish Solanki

I help organizations unlock the power of Artificial Intelligence by designing and deploying intelligent systems that automate workflows, analyze complex data, and generate real-world impact. With a Master’s in Artificial Intelligence and hands-on experience across NLP, Computer Vision, and Large Language Models (LLMs), I specialize in building robust, end-to-end AI solutions using cutting-edge tools like LangChain, LangGraph, and Retrieval-Augmented Generation (RAG) frameworks. My work spans the full AI lifecycle — from model development to deployment — and includes- Pre-training and fine-tuning LLMs (Granite, LLaMA) to enhance reasoning and performance - Building LLMs from scratch on custom domain-specific data - Designing agentic AI systems using multi-agent coordination for intelligent task automation - Developing RAG-powered chatbots and knowledge retrieval systems with FAISS, ChromaDB, and MongoDB - Inferencing and fine-tuning vision models (Qwen, granite-vision, Nanonets-OCR) for tasks like document intelligence and OCR - Optimizing inference speed using VLLM and VLLM with Docker for efficient, scalable deployment of LLMs and vision models - Training YOLO models for visual detection tasks - Creating real-time vision tools with OpenCV and MediaPipe for gesture-based interaction and video analysis - Automating data annotation, extraction, and cleaning pipelines — reducing manual effort by over 70%- Web scraping and task automation using Python and Selenium to streamline data gathering and repetitive processes - NLP-driven pipelines including transcription, topic modeling, semantic search, and sentiment analysis - Designing web apps with Flask and Streamlit for data preparation, model interaction, and live inference dashboards - Researching Quantum Machine Learning, with award-winning work on Quantum-GANs for high-resolution fashion image generation (Best Paper @ TIACOMP 2024) Core Skills & Tools: LLMs (GPT, Claude, Granite, LLaMA)| Vision Models (Qwen, Granite, NanonetOCR)| VLLM + Docker | LangChain | LangGraph | RAG | OpenCV | MediaPipe | Python | Flask | Streamlit | MongoDB | Selenium | Quantum GANs | Reinforcement Learning | NLP | Data Science | Deployment Automation What sets me apart is my ability to combine deep AI research with real-world engineering — building scalable, production-ready AI systems that drive tangible business outcomes. Let’s connect and explore how AI can power the next generation of your products, services, or internal tools.