{"$schema":"https://raw.githubusercontent.com/jsonresume/resume-schema/v1.0.0/schema.json","basics":{"name":"Sangam Gupta","label":"AI/ML engineer","email":"contact@sangamgupta.in","phone":"+91 9970116435","url":"https://www.sangamgupta.in","summary":"Sangam Gupta is an ML engineer in Mumbai who builds models in Python and scikit-learn and ships AI products, including an Android app with 500+ downloads.","location":{"city":"Mumbai","countryCode":"IN"},"profiles":[{"network":"GitHub","url":"https://github.com/TheSangamX"},{"network":"LinkedIn","url":"https://www.linkedin.com/in/thesangamx"},{"network":"X","url":"https://x.com/TheSangamX"},{"network":"Medium","url":"https://medium.com/@TheSangamX"},{"network":"DEV","url":"https://dev.to/thesangamx"},{"network":"Hugging Face","url":"https://huggingface.co/TheSangamX"}]},"work":[{"name":"Self-directed","position":"Independent AI/ML engineer","summary":"Training, evaluating and serving ML models (Customer Intelligence, Credit Risk, NewsIntel), and shipping AI products with AI coding assistants: ProfilePilot (500+ downloads on Google Play) and DealShield AI (Chrome Web Store)."},{"name":"IT Vedant","position":"Data science intern","summary":"Real-world ML projects in business insight, predictive modelling and data storytelling."},{"name":"JPMorgan Chase","position":"Quantitative research","summary":"Financial data modelling and statistical research on market data."},{"name":"British Airways","position":"Data science","summary":"Predictive modelling of customer behaviour and an end-to-end ML pipeline."},{"name":"Deloitte Australia","position":"Data analytics","summary":"Data cleaning, visualisation and business insight for client deliverables."},{"name":"Tata","position":"Data visualisation and GenAI","summary":"Power BI dashboards and business reporting."},{"name":"Quantium","position":"Retail analytics","summary":"Customer analytics and statistical analysis for the grocery sector."}],"education":[{"institution":"CSMIT, University of Mumbai","studyType":"B.E.","area":"Computer Engineering","startDate":"2022-06","endDate":"2026-05","score":"7.25 CGPA"}],"certificates":[{"name":"Professional Certification, Data Science and Analytics with AI","issuer":"IT Vedant","date":"December 2025 to 2026"},{"name":"Machine Learning with Python","issuer":"IBM, via IT Vedant","url":"https://courses.itvedant.skillsnetwork.site/certificates/0b305cd571dc4d1593afa2884f8dfe1e"},{"name":"Data Analysis with Python","issuer":"IBM, via IT Vedant","url":"https://courses.itvedant.skillsnetwork.site/certificates/093d0849059048259e69eb82e9ca8316"},{"name":"Python 101 for Data Science","issuer":"IBM, via IT Vedant","url":"https://courses.itvedant.skillsnetwork.site/certificates/f9f0f0b1a9184ec4826eec29697aa9be"},{"name":"Advanced Excel, SQL, Power BI and Python for Data Analytics","issuer":"IT Vedant"},{"name":"Machine Learning and NLP Bootcamp","issuer":"Udemy","date":"2025"},{"name":"Stock Market Course","issuer":"Asmita Patel Global School of Trading","date":"2024"}],"skills":[{"name":"Languages and data","keywords":["Python","SQL","pandas","NumPy","Statistics","EDA","Matplotlib","Seaborn","Jupyter"]},{"name":"Machine learning","keywords":["scikit-learn","Feature engineering","Model evaluation","K-Means","PCA","RFM analysis","NLP (TF-IDF, NLTK)"]},{"name":"Serving and tools","keywords":["FastAPI","Streamlit","AWS EC2","Git","GitHub"]},{"name":"Analytics","keywords":["Power BI","Excel"]},{"name":"AI-assisted development","keywords":["Claude Code","Cursor","Copilot","ChatGPT","Gemini","Specs and testing","Vercel and Render deploys","Play Store and Chrome Web Store launches"]}],"projects":[{"name":"ProfilePilot","description":"An Android app with 500+ downloads that coaches people to a stronger professional profile, without inventing a single achievement.","highlights":["AI profile coach on Google Play: profile score and audit, headlines, About, experience bullets, job match, messages, interview prep and posts. Eight tools behind one Supabase Edge Function calling Groq's gpt-oss-120b with strict JSON schemas.","500+ downloads on Google Play","8 AI tools, each returning a strict JSON schema","0 data collected, per the Play Store data safety listing"],"keywords":["Android","Supabase Edge Functions","Deno","Postgres","Groq","gpt-oss-120b"],"url":"https://play.google.com/store/apps/details?id=in.sangamgupta.linkedinoptimizer"},{"name":"DealShield AI","description":"A Chrome extension that reads a tough negotiation email in Gmail and drafts three fact-checked replies you can insert in one click.","highlights":["Negotiation copilot for Gmail, published on the Chrome Web Store. A React extension reads the open thread, an Express API asks Groq's gpt-oss-120b for three strategies under a strict JSON schema, and a second pass checks every draft against the email. Free and Pro plans, with Razorpay subscriptions.","3 reply strategies per email, each with the reasoning behind it","2 AI passes: draft, then fact-check against the email","10 free analyses a month; Pro is unlimited"],"keywords":["Chrome Extension (MV3)","React","Vite","Tailwind CSS","Node.js","Express","Zod","Groq","gpt-oss-120b","Supabase","Razorpay","Render"],"url":"https://chromewebstore.google.com/detail/kbpklkipafiimileagddpmbljlmgofia"},{"name":"sangamgupta.in","description":"This site: a portfolio built as an AI system, with an agent that answers from my own content and the testing and monitoring around it in public.","highlights":["Next.js site whose Ask Sangam agent retrieves passages from the site with BM25 plus Gemini embeddings fused by reciprocal rank fusion, answers through Groq with Gemini as backup, and falls back to answering from the content with no model. Weekly evals, a red-team page, provider telemetry, drift monitoring for the demo models and an MCP server.","45 test questions the agent answers through the live endpoint every week","9/10 reworded questions retrieved correctly by the hybrid retriever, vs 8/10 for keywords alone","3 ways to answer: Groq, then Gemini, then the site's own facts with no model"],"keywords":["Next.js","TypeScript","Groq","Gemini","BM25","Embeddings","Upstash Redis","MediaPipe","WebLLM","Vercel","GitHub Actions"],"url":"https://www.sangamgupta.in/work/sangamgupta-in"},{"name":"Quanta (Sangam AI)","description":"A ChatGPT-style assistant running on my own server, set up with AI help.","highlights":["Self-hosted LLM chat platform running Qwen2.5 through Ollama and OpenWebUI, deployed on DigitalOcean behind Nginx with HTTPS and custom system prompts."],"keywords":["Qwen2.5","Ollama","OpenWebUI","Nginx","DigitalOcean"],"url":"https://ai.sangamgupta.in"},{"name":"NewsIntel","description":"Paste a headline or article and get a real-or-fake verdict in seconds.","highlights":["My final-year engineering project: an NLP platform that classifies news as real or fake with TF-IDF features, NLTK preprocessing and a Passive Aggressive classifier, served through Flask with Google sign-in and verification history on Supabase."],"keywords":["Python","scikit-learn","TF-IDF","NLTK","Flask","Supabase","Google OAuth"],"url":"https://newsintel.sangamgupta.in/login"},{"name":"Customer Intelligence","description":"Turns raw transactions into customer segments a business can act on, served to a web app and an Android app from one API.","highlights":["End-to-end unsupervised ML system: RFM features and K-Means clustering behind a FastAPI service on AWS EC2, used by a Streamlit web app and a native Android app."],"keywords":["Python","scikit-learn","K-Means","FastAPI","AWS EC2","Streamlit","Kotlin"],"url":"https://customer-intelligence-segmentation.streamlit.app/"},{"name":"Credit Risk Predictor","description":"Estimates how likely a loan applicant is to default, tuned to catch more risky applicants rather than to look accurate.","highlights":["Gradient-boosted model on applicant financial, demographic and credit data, with a Streamlit app that returns default probability, risk class and the features driving the model.","0.76 ROC-AUC","41.8% of defaulters caught (recall) at a 15% threshold","85.2% accuracy, which hides how rare defaults are"],"keywords":["Python","scikit-learn","HistGradientBoosting","pandas","Streamlit"],"url":"https://huggingface.co/spaces/TheSangamX/credit-risk-predictor"},{"name":"LinkVault","description":"Save the links you want to keep, sort them into categories and find them again with search.","highlights":["Personal link manager for Android, published on Google Play: add a URL with an optional title and note, file it under a colour-coded category, search everything, and switch to dark mode. The Play Store data safety listing declares no data collected or shared."],"keywords":["Android"],"url":"https://play.google.com/store/apps/details?id=in.sangamgupta.linkvault"},{"name":"AI & ML RAG Assistant","description":"Ask questions about a PDF and get answers with the page numbers they came from.","highlights":["Learning prototype of retrieval-augmented generation: LangChain chunking, MiniLM embeddings in a FAISS index, top-3 retrieval, and Gemini via OpenRouter answering only from the retrieved context."],"keywords":["Python","LangChain","FAISS","sentence-transformers","Streamlit"],"url":"https://github.com/TheSangamX/ai-ml-rag-assistant"},{"name":"Student Marks Predictor","description":"Predicts a student's final exam marks from 18 academic, study and lifestyle factors.","highlights":["scikit-learn regression pipeline served from a FastAPI backend on AWS EC2 to a Streamlit app and an Android app in closed testing on Google Play (API design and Android app built with AI assistance)."],"keywords":["Python","scikit-learn","FastAPI","Pydantic","AWS EC2","Streamlit"],"url":"https://thesangamx-student-marks-predictor.streamlit.app/"},{"name":"India Election 2024 Dashboard","description":"Explore India's 2024 general election results by state, demographics and constituency.","highlights":["Power BI dashboard with a data model for state-wise and constituency-level analysis, across four views: overview, state demographics, political landscape and constituency analysis."],"keywords":["Power BI","Data modelling"],"url":"https://generalelection2024.vercel.app"},{"name":"AI Signal","description":"Daily stock signals, scored on RSI and sent straight to Telegram.","highlights":["Automated screener that pulls market data with yfinance, scores stocks on RSI-based indicators and sends daily signals through a Telegram bot."],"keywords":["Python","yfinance","pandas","Telegram Bot API"],"url":"https://github.com/TheSangamX"},{"name":"Smart Expense Categorizer","description":"Turns a list of transactions into categories and spending insights.","highlights":["Auto-categorises transactions and surfaces spending insights in a Streamlit app."],"keywords":["Python","pandas","Streamlit"],"url":"https://github.com/TheSangamX"},{"name":"House Price Prediction","description":"Estimates a California district's median house value; the best of nine models I compared explains 83% of the variance.","highlights":["Nine regression models compared on the California housing data (20,640 census block groups), from linear baselines to tree ensembles. The winner, a HistGradientBoostingRegressor, now runs in the browser on this page.","0.835 R² on held-out data","$46.5k RMSE","9 models compared"],"keywords":["Python","scikit-learn","pandas","HistGradientBoosting"],"url":"https://github.com/TheSangamX/California-Housing-Price-Prediction-End-to-End-ML-Project"},{"name":"Mobile Recommendation System","description":"Shortlists smartphones from scraped GSM Arena specs.","highlights":["Scrapes GSM Arena with BeautifulSoup, cleans the data with pandas and serves recommendations in Streamlit."],"keywords":["BeautifulSoup","pandas","Streamlit"],"url":"https://github.com/TheSangamX"}],"meta":{"canonical":"https://www.sangamgupta.in/resume.json","lastModified":"2026-11-13"}}