Investment Research
DualLens: AI powered investment research
Ranks companies on how their shares have done and on what they say about their AI plans, in one step.
01The problem
An analyst who wants companies that are both financially strong and ready for AI has to work in two places. Share performance lives in market data. Strategy lives in long company documents. Putting the two side by side means hours of reading and copying between them.
02What it does
DualLens joins both views. It takes three years of share performance and combines it with what each company’s own strategy documents say about adopting AI. The result is one ranked list showing which companies look strong on both counts.
03How it works
The system first reads the strategy documents and stores them so it can search them by meaning, not just by keyword. When asked a question, it finds the most relevant passages, pulls the share data, and brings both together in its answer. Every insight points back to the document it came from.
04What came out of it
One question now does what used to be a manual loop of searching, reading and cross checking. It also showed me how much of an analyst’s time goes into gathering rather than thinking.
Built with
Want the technical detail?
The code and the full technical write up are on GitHub.