Portfolio Lens

Inspiration

What it does

How we built it

Challenges we ran into

Accomplishments that we're proud of

What we learned

What's next for Portfolio Lens

What it does

Portfolio Lens makes the structure of an ETF portfolio legible before a change is proposed. It opens with a Portfolio X-ray: direct positions, look-through exposure, sector concentration, and risk on one fixed 36-month path.

Why the sample is synthetic

Data-source redistribution terms prevent packaging and distributing a real ETF constituent dataset for this demonstration. Every basket, security, holding, and return path shown here is fictional and visibly labelled as simulated. This is not investment advice, a recommendation, forecast, or order path.

How it works

After explicit consent, GPT-5.6 turns a plain-language goal into a bounded, reviewable policy draft. It may name only the portfolio's selected ETF universe and cannot add a ticker, calculate results, or bypass reviewer acceptance. After acceptance, the deterministic engine searches simulated scenarios and displays the proposed trade, risk comparison, constraint checks, and reproducibility identifiers together.

Built with Codex and GPT-5.6

Codex accelerated the plan-build-test-review loop, helping break the product boundary into verifiable units and maintain the release evidence. GPT-5.6 has one narrow, server-side role: interpreting consented intent. All portfolio calculations and numerical comparisons remain deterministic, and the browser never receives an OpenAI API key.

Built With

• codex