WORKING AI PRODUCTS

Functional prototypes for pricing intelligence and quantitative decision support, built to expose the evidence behind every recommendation.

Built with Next.js, validated data pipelines, transparent calculations, portfolio-hosted APIs, and clearly stated product limits.

Featured product build

MarginLens

A pricing-intelligence workspace that validates sales data, ranks account-level margin shortfalls, exposes the supporting invoices, and tests price-only target scenarios without hiding the assumptions.

Interactive case study
01

Validated input

Rejects incomplete or malformed sales data.

02

Auditable math

Shows revenue, cost, profit, and source rows.

03

Scenario testing

Separates benchmark gaps from pricing uplift.

Try the portfolio demoSample data · no sign-in required

Featured Build

The live widget below runs through this portfolio's API route, so it works on the hosted site without a separate Python server.

Live quant widget

Stock Signal Predictor

A live stock-direction experiment that fetches recent market data, computes technical indicators, and returns a transparent next-session signal.

Market DataTechnical IndicatorsBacktestNext.js APILive WidgetNot Financial Advice
Enter a ticker to run the live signal.

The result includes direction, confidence, a simple time-ordered backtest, RSI, MACD, and last close.

Statistical estimate only. This page is not investment, trading, or financial advice.

© 2026 James Boggs. Machine Learning Engineer.

Florida · Remote-first · Working AI tools in Python, Next.js, and Vercel