Headline

AI/ML engineer @Express analytics||Ex-Data science intern@Rupeeflo||Ex-analyst intern@Signzy||Ex-Data science intern@slice||Indian Institute of Information Technology

About

Data Scientist | Machine Learning Engineer | Generative AI I am a AI/ML Engineer with 1.5 years experience in building AI solutions across fintech and enterprise domains while working with slice,signzy , Rupeeflo and express analytics,I have worked on Document AI/OCR, Retrieval Augmented Generation (RAG), NLP, speech AI, data analytics, and intelligent automation, delivering scalable solutions for real world business challenges. My expertise includes Machine Learning, Deep Learning, Generative AI, MLOps, and NLP, with hands-on experience in Python, SQL, PyTorch, TensorFlow, LangChain, LangGraph, LlamaIndex, FastAPI, Docker, MLflow, and modern LLM ecosystems. I enjoy designing end-to-end AI systems, from data processing and model development to deployment and production. Passionate about applying AI to solve complex problems, I am always interested in building intelligent, reliable, and impactful solutions that create measurable business value.

Location

Sangli

AI stack

anthropic apis

codex

cursor

hugging face

langgraph

langsmith

Work experience

AI/ML engineer at Express Analytics, Jun 2026 - Present · 4 mos, California, United States

Data Science Intern at Rupeeflo - NRI Banking & Wealth Platform, Jan 2026 - Jun 2026 · 6 mos, Bengaluru, Karnataka, India, •Passport Verification System Built an end-to-end automated Passport Verification System to eliminate manual verification overhead and reduce processing time. Leveraged Ollama (Qwen3-VL:8B) for OCR and structured data extraction with custom schema enforcement, and implemented multi-layer validation (checksum + rule-based checks) to detect tampered or incorrect documents. Achieved 78% extraction accuracy on 300 samples and 100% precision (0% false positives) in flagging invalid passports. • Bank Statement Analyzer Built an end-to-end bank statement processing pipeline to automate loan decisioning workflows. Engineered extraction and normalization of raw financial data from PDFs, applied multi-check integrity validation across balances, transaction chains, and date ordering, and generated structured Excel reports with AI-driven underwriting flags and risk signals

Analyst at Signzy, Nov 2025 - Jan 2026 · 3 mos, Bengaluru, Karnataka, India, • Analyzed KYB datasets to support business verification workflows, ensuring data accuracy and consistency across multiple sources. • Worked on IEC and EPFO data analysis, validating records and assisting in structured financial and organizational verification processes. • Contributed to the Auto Tracking product by writing Python scripts to collect, clean, and validate vehicle related data from various sources. • Performed data filtering, validation, and organization to improve data quality and reliability for downstream analytics and decision making. • Collaborated with internal teams to understand data requirements and deliver well structured, analysis ready datasets.

Data Science Intern at slice, Feb 2025 - Aug 2025 · 7 mos, Bangalore, • Built and deployed an end to end LegalBot for querying RBI regulations using LlamaIndex, FAISS, and Azure OpenAI, enabling accurate, context aware responses for legal and compliance teams. Built automated pipelines for RBI document scraping, multi source ingestion (user uploads & scheduled backend updates), and metadata aware retrieval, and developed a Streamlit UI for real-time search and LLM powered Q&A, integrating ingestion, retrieval, and inference into a production ready system. •Engineered and optimized LLM prompts to enhance conversational tone, improve response quality, and embed a custom persona into the chatbot for a more engaging and user aligned interaction experience. • Conducted PoC for the Voice pay product of slice , benchmarking open-source and proprietary STT models using WER, BLEU score, and latency metrics to determine the optimal solution. Recommended deployment of the selected model, which delivered a 25% improvement in transcription accuracy compared to the existing system

Contributor at GirlScript Summer of Code, Oct 2024 - Nov 2024 · 2 mos

Education

Bachelor of Technology - BTech, Indian Institute of Information Technology Senapati, Manipur

12th., sanjay ghodawat institute

Public links

github: https://github.com/pratikkodag