Nicholas Belous

Engineering a trade-event data pipeline for FBSI 2.0, Fidelity's brokerage modernization. The platform processes 3.5M+ daily trades and fans out to 5+ downstream analytics consumers.

Built automated pipelines scraping 13+ sources across 18,000+ companies and validating 100,000+ financial records, surfacing performance and sentiment signals that informed institutional trading strategies.

Built a Python NLP pipeline (LDA topic modeling + NLTK sentiment) over 500+ executive survey responses and produced 30+ visualizations to support client go-to-market strategy.

Develop go-to-market strategy and growth analytics for early-stage startups, segmenting users from survey data and designing Google Ads / SEO campaigns against measurable ROI targets.

A Chrome extension that scrapes Facebook Marketplace listings and scores them in real time via the Claude API, returning a 0-100 deal score, a used-market price range, and a suggested offer without leaving the page.

A full-stack quant-finance analyzer ingesting financials and price history for 500+ S&P 500 firms, computing annualized return, volatility, beta, and Sharpe ratio, plus a ridge-regression model benchmarked against a naive baseline.

A C++ backtracking solver for the LinkedIn Queens puzzle, using constraint propagation over an N-Queens-style board.

A custom chess engine using alpha-beta pruning and evaluation heuristics, a nod to the game that started my love of problem-solving.

An automated internship-sourcing system that scrapes and standardizes postings across 50+ job boards into a structured database.