Ajay Rajaram

Background

About Ajay Rajaram

AI/ML Engineer | LLM GenAI & NLP Engineer | Data Science Professional With around 4 years of professional experience across industry and government I specialize in building production-grade AI solutions—from scalable document intelligence systems to LLM-powered search and conversational interfaces. At the Maryland Motor Vehicle Administration (MDOT MVA) I lead the design and deployment of cutting-edge AI models that automate document processing enable secure and scalable vector search with RAG architectures and power real-time chatbots using frameworks like Rasa and LangChain. I’ve built full-stack AI pipelines using LangChain Ollama FAISS PyTorch Azure Vision and custom NER models with SpaCy—deployed via Flask IIS and WFastCGI for enterprise-grade reliability. My work in vector-based retrieval OCR LLM fine-tuning and hybrid search has improved information access and automation in high-volume environments. Previously at PwC I focused on data analytics and NLP prototyping for Fortune 500 clients combining ETL Power BI and statistical modeling. I’m also skilled in optimizing ML pipelines using multiprocessing parallel computing and CUDA-based acceleration for high-throughput use cases. ⚙️ Skills: Python PyTorch LangChain FAISS Rasa Azure Vision OCR (Tesseract PaddleOCR) Flask SQL JavaScript SpaCy BERT Sentence Transformers Ollama 📌 Expertise: LLMs RAG Systems NLP Document AI OCR Chatbots Custom NER AI-Powered Search 🎓 Education: MPS – Data Science University of Maryland Baltimore County (UMBC) | B.Tech – Information Technology GRIET India 🚀 Passionate about using AI to simplify complex systems empower users and build human-centric solutions. Let’s connect and innovate together! Reach me at: ajay.rajaram99@gmail.com