Pritish Das

Results-driven Data & AI Engineer with 3.8 years of experience specializing in building production-ready AI solutions, scalable data pipelines, and backend architectures. Expert in Generative AI Architecture (Gemini, OpenAI, Claude) and FastAPI. Experienced in Prompt Engineering, Google ADK, and designing complex LLM workflows and prompt-based automation systems that transform business requirements into high-performance applications.

Python Developer - Pichain Innovation Ltd

(2022-07)

• Gen AI Architecture: Designed and deployed advanced LLM workflows using Google Gemini and OpenAI for automated summarization and entity extraction, reducing manual data processing by 70%.
• Voice-Bot Orchestration: Optimized conversational AI flows and backend orchestration using Ultravox, improving call transcription accuracy and reducing response latency for real-time voice agents.
• Deployment & Infrastructure: Managed code deployment on GCP VMs and utilized Linux environments for backend orchestration; handled data storage and insertion using GCP Buckets.
• API Development: Designed and documented high-performance RESTful APIs using FastAPI, integrating with MongoDB to support scalable data storage for 10k+ records.
• Web Automation: Engineered robust data scraping bots using Selenium, BeautifulSoup and Playwright, automating extraction from complex web architectures with a 95% success rate.
• Computer Vision: Developed and fine-tuned YOLOv5 models to automate signature extraction from legal and financial documents, increasing processing speed by 60%.

Data Analyst Intern - Coding Ninjas

(2022-01 - 2022-06)

• Business Intelligence: Leveraged data from e-commerce platforms (Shopify, Shipmonk, etc) to generate comprehensive business insight reports for senior developer teams.
• Data Acquisition: Developed web scraping scripts to aggregate market data from multiple sources, enhancing the internal database for predictive modeling.

Bachelor of Technology - B.Tech - Gandhi Institute For Technology, BPUT, Odisha, India (2020)