Senior Data Scientist
Khushbu is a skilled NLP and Generative AI developer with 8 years of deep expertise in LLMs, RAG pipelines, and prompt engineering. She builds intelligent applications using models like BERT, CLIP, and Meta-LLaMA, integrated via LangChain and FastAPI. From sentiment analysis to research chatbots, Khushbu delivers scalable AI solutions with strong Python, PostgreSQL, and Docker skills, turning complex ML challenges into powerful impact.
Julia
4 years
Akka
2 years
Designed and developed a Chrome extension for uploading academic PDFs and enabling conversational Q&A. Built FastAPI backend services for parsing documents, managing chat sessions, and integrating LLMs for intelligent responses. Implemented document chunking, embedding with Hugging Face models, and Retrieval-Augmented Generation (RAG) using ChromaDB for accurate, context-aware answers. Ensured system reliability, responsiveness, and scalability across research queries and document comparison workflows.
Developed an AI-driven document verification system to validate user-filled form data against structured datasets. Extracted text from scanned PDF forms using OCR and compared it with reference Excel data using word and character error rate (WER/CER) metrics. Designed matching logic with configurable thresholds to automate acceptance or rejection. Streamlined the bank's verification process by eliminating manual review through intelligent automation and format normalization.
Enhanced an intelligent job matching platform by optimizing resume-to-job description alignment using advanced NLP and GenAI techniques. Researched and integrated multiple embedding models to improve contextual understanding and match accuracy. Leveraged LLMs to extract key information from resumes and job posts, refining the core matching algorithm. Conducted performance benchmarking, extensive testing, and debugging to ensure reliable, accurate, and scalable matching functionality within the recruitment platform.
CHARUSAT University
Bachelor of Technology in Computer Science & Engineering

Khushbu’s work on our clinical NLP pipeline was exceptional. Her ability to transform unstructured patient data into structured insights significantly improved our reporting accuracy. She combines technical depth with a strong understanding of real-world application, which made her a valuable partner in our AI initiatives.

We brought Khushbu on board to enhance our job matching algorithm, and the results were beyond impressive. She implemented advanced embeddings and LLM-driven extraction that increased match relevance by over 35%. Her attention to detail and commitment to quality set her apart.

Khushbu led the development of our automated document verification module. Her precision in aligning OCR output with validation logic was crucial to improving efficiency and reducing errors. She communicates clearly, delivers reliably, and is always ready to adapt to evolving project needs.

Our conversational research assistant wouldn’t exist without Khushbu’s innovation. From embedding logic to RAG pipeline design, she took care of every technical aspect while ensuring a seamless user experience. Her blend of ML knowledge and product thinking was exactly what we needed.