Headline

AI & Machine Learning Engineer | LLM Fine-Tuning | NLP, RAG Pipelines & Scalable AI Solutions

About

I don’t just build models — I build intelligent solutions that make businesses smarter. From fine-tuning LLMs like GPT & LLaMA to crafting RAG pipelines, semantic search engines, and chatbots that actually feel human — I turn AI buzzwords into working products. My sweet spot? NLP/NLU systems that understand context Scalable ML pipelines that run like clockwork AWS-powered deployments that just work I speak fluent PyTorch, TensorFlow, FastAPI, Docker, AWS — and I’m always experimenting with the next big thing in AI. If it’s innovative, data-driven, and solves a real problem, I’m in. Let’s build something game-changing.

Location

Greater Bengaluru Area

AI stack

anthropic apis

claude code

claude

codex

gemini

langchain

Work experience

Data Scientist at Enverus, Nov 2025 - Present · 1 yr, Bengaluru, Karnataka, India

Software Engineer at MPS Limited, Aug 2023 - Nov 2025 · 2 yrs 4 mos, Bengaluru, Karnataka, India, AI-Powered Client Support Chatbot (RASA) Reduced human agent intervention by 40% and cut average resolution time by 1 business day by designing and deploying an AI-powered client support chatbot on the RASA framework. The bot delivers relevant query recommendations and autocompletion — matched using BLEU score similarity — to guide users toward faster resolutions, and automatically generates support tickets for any query it can't resolve, removing manual triage from the workflow entirely. Technical Support AI Agent — Fine-Tuned LLaMA 2 + RAG Built a specialized technical support agent by fine-tuning LLaMA 2 (7B) on our internal platform (DigiEdit) to generate comprehensive, context-grounded technical resolutions — going beyond generic chatbot responses to platform-specific troubleshooting. Built a retrieval-augmented generation (RAG) pipeline so answers stay grounded in real documentation: FAISS vector store for retrieval, HuggingFace embeddings, and LangChain's RetrievalQA chain with custom prompt templates. Used CTransformers for efficient local inference, and built the interface on Chainlit's async framework for responsive, real-time conversations

Junior Software Engineer -ML at entomo, Jun 2023 - Aug 2023 · 3 mos, Bengaluru, Karnataka, India, • Spearheaded global web scraping initiatives, efficiently extracting structured data from various ATS portals, job listing platforms, and social media channels such as LinkedIn, Twitter, and Facebook across multiple geographies. • Automated end-to-end data collection pipelines, optimizing workflows, enhancing scalability, and reducing manual intervention. • Developed and maintained an interactive dashboard to visualize daily data extraction, enabling real-time insights and decision-making. • Led the design, development, and optimization of a Named Entity Recognition (NER) model, accurately identifying key entities and phrases from job descriptions to enrich extracted datasets. • Collaborated with cross-functional teams to validate and fine-tune the NER model, improving extraction precision and enhancing the overall efficiency of the data pipeline.

Education

Bachelor of Engineering - BE, Visvesvaraya Technological University

Public links

medium: https://medium.com/@ghkarthik4