Most AI strategy tools sound confident even when they're guessing. AnchorStrategy turns your meeting notes into a strategy report where every claim — every number, every company name — is checked against the actual source sentence it came from. If the evidence isn't there, it says "insufficient basis" instead of inventing a figure.
I kept running into the same problem: AI strategy tools sound completely confident even when they're guessing. I once fed one a set of meeting notes with zero hard numbers in them, just to test it — it came back with a confident "$400K in projected savings." Made up entirely.
That's what pushed me to build AnchorStrategy. Every claim in the generated report has to trace back to the actual sentence it's supposedly based on — company names, numbers, timeframes all have to match the source. If they don't, it flags "insufficient basis" instead of stating it as fact.
It's still early (solo build, private beta), and I'm actively hardening the citation-fidelity checks based on real edge cases people find. Would genuinely love feedback, especially if you've been burned by an AI tool inventing numbers before.