Launching today

Walmart Business 360

Launching today

From retail forecasts to inventory and FP&A decisions

1 follower

From retail forecasts to inventory and FP&A decisions

1 follower

Visit website

Walmart Business 360 turns retail forecasting into decision-ready finance and operations evidence. Explore demand forecasts, inventory cost trade-offs, FP&A scenarios, an auditable Excel model, and an interactive Power BI report—all backed by governed, reproducible artifacts. Built as an independent portfolio project using public M5, SEC, and FRED sources; not affiliated with Walmart.

Free

Launch tags:Analytics•GitHub•Finance

Launch Team

OpenAI

GPT-6 Astra Challenge

How did Astra change the scope or ambition of what you built?

Maker

GPT-6 Astra expanded this from a forecasting notebook into an end-to-end, auditable decision product. It helped me inspect and repair a large codebase, trace governed execution failures, strengthen tests and provenance controls, connect forecasts to inventory and FP&A decisions, and turn the results into a verified Excel model, a portable Power BI report, a public Next.js site, GitHub release, and production deployment. Astra also helped preserve an important analytical boundary: the simpler Seasonal Naive S7 won the held-out WRMSSE benchmark, so the project reports that result instead of forcing the more complex model to look better. The deployed site does not call Astra at runtime; Astra was the development, verification, and release collaborator that made this broader, production-quality scope possible.

Report

2h ago

Maker

📌

Hi Product Hunt! I’m Vicenzo, an Economics and Finance graduate who built Walmart Business 360 to connect forecasting with decisions people in FP&A and operations can actually inspect. The project moves from M5 demand forecasts to inventory trade-offs, scenario analysis, an Excel financial model, and a four-page Power BI report. The public site also exposes the methodology, limits, downloads, and reproducibility evidence. GPT-6 Astra helped me work end to end across the codebase: audit the analytical pipeline, trace failures, strengthen tests and governance, refine the public narrative, and prepare the deployment. The live site itself does not call Astra at runtime; Astra was the development and verification collaborator. One result I kept intentionally honest: the more complex Candidate C did not win the held-out WRMSSE benchmark—Seasonal Naive S7 did. That transparency is part of the product. I’d love feedback on whether the story makes the forecast-to-finance trade-offs clear and useful.

Report

7d ago

OpenAI

GPT-6 Astra Challenge

How did Astra change the scope or ambition of what you built?

Maker

Report

2h ago

Maker

📌

Report

7d ago