Vicky Kumar

I’m Vicky Kumar, an AI Engineer and Data Scientist currently pursuing a B.S. in Data Science and Applications at IIT Madras. My core expertise lies in Large Language Model (LLM) training, fine-tuning, and evaluation, combined with a deep foundation in Python development, FastAPI-based automation, and full-stack AI integration.

Through my professional experience with Outlier and Soul AI, I’ve contributed to large-scale AI training and evaluation initiatives, improving the reasoning, factual accuracy, and contextual understanding of modern language models. At Outlier, I trained models in Python programming, algorithmic reasoning, and mathematical problem-solving, while designing persona-driven SFT datasets and evaluating multi-turn conversational quality. At Soul AI, I focused on AI response evaluation, ensuring alignment, coherence, and factual reliability across diverse domains. These experiences refined my expertise in prompt engineering, RLHF workflows, model assessment, and data quality optimization.

Beyond model training, I’ve built a series of intelligent systems and data-driven platforms, such as the AI Engineer Portfolio Chatbot, Vicky Data Science Assistant, and Chatbot for TDS, integrating LangChain, FastAPI, Docker, and vector databases (FAISS/Chroma) for advanced reasoning and automation. My recent project, PolyMandi, reflects my ability to architect scalable, production-ready systems that merge AI reasoning with real-world business functionality.

I have strong command over technologies like TensorFlow, Scikit-learn, Pandas, NumPy, Hugging Face, and Azure, along with deep experience in model fine-tuning, reinforcement learning, and RAG pipelines. I thrive at the intersection of data engineering and AI innovation, where precision and creativity combine to deliver measurable impact.

I’m passionate about developing AI systems that are trustworthy, transparent, and truly intelligent, capable of learning beyond static datasets. My goal is to advance the frontiers of AI automation, model interpretability, and human-aligned intelligence, while continuing to grow as a researcher and engineer.

I value opportunities that offer both intellectual growth and fair financial reward, enabling me to continue creating meaningful, scalable AI systems that push the boundaries of what’s possible.