Soham Kukreti

CS Student  /  Engineer  /  Builder

About

Hi, I am Soham, a 22-year-old Computer Science student at JIIT Noida. I love building cool stuff using code, with Python being my language of choice. I am currently working at Crawl4AI, where we are democratising data with our open source web crawler and cloud API. Additionally, I produce music, play the guitar, and beatbox.

Experience

  1. Engineer — Crawl4AI Dec 2025 – Present
    • Developing the Crawl4AI n8n community node, enabling seamless workflow automation and integrations.
    • Fixing bugs and improving stability in the open-source Crawl4AI codebase, contributing to ongoing product reliability and performance.
    • Performing testing, providing technical guidance, and assisting with onboarding new users to the Crawl4AI Cloud.
  2. Data Science Intern — MagicBricks Jun 2025 – Jul 2025
    • Engineered a RAG pipeline for real estate PDFs using Tesseract OCR, Google Embeddings API, and Weaviate.
    • Processed 20,000+ properties to find geospatial matches within 3–5 km for a RAG-based property recommendation bot.
    • Applied scalar quantization to compress vector embeddings, reducing memory footprint by 75%.
  3. Intern — jobup.ai Aug 2024 – Nov 2024
    • Built backend APIs with Flask and tested them using Postman.
    • Developed user-centric applications leveraging large language models.
    • Implemented advanced LLM techniques such as RAG and Structured Outputs to build dynamic systems.
  4. Technical Coordinator — OSDC, JIIT Noida Jul 2024 – Present
    • Promoted open-source culture within the college.
    • Organised hackathons, events, and talks engaging over 800 students.

Projects

  1. URLias — A cross-browser extension that lets you jump to your favourite sites using short aliases, supports wildcard matching, omnibox integration, and sync storage.
  2. Omilia — An educational platformer game developed to help children learn new languages through an engaging format.
  3. BCI Gaming — A project aimed at allowing people to play video games using EEG signals generated from the brain.
  4. FinWiser — A financial analysis tool featuring real-time market data, financial tracking, SIP investment planning, and interactive charts. Built with Python, Plotly, Streamlit, NumPy, and Pandas.
  5. SignSense — A sign language detection application developed using OpenCV, MediaPipe, TensorFlow, and Pygame.
  6. Todo Tasks Tracker — A task management application developed with Django and hosted on Railway with a PostgreSQL database.
  7. ResumeAnalyzer — A tool that analyses resumes against job descriptions, providing match scores, missing keywords, and personalised suggestions. Built with Flask and Python.
  8. miprofilio — A unified dashboard that aggregates coding profiles from various platforms, extracts relevant skills, and displays them in a consolidated view. Developed using TypeScript and Python.

Achievements

Skills