I stopped treating more advice as the answer to hard decisions

I kept noticing the same failure in my own work: difficult decisions were rarely blocked by a complete lack of advice. They were blocked by one confident answer arriving with no visible disagreement, no assumptions, and no record of why I chose the move.

That became the reason I started building SUDHI.

The first version tried to do too much. It sounded like a productivity system and asked people to fit their decision into its language. The useful breakthrough was smaller: make the user the CEO, put one real decision on the table, and give ten fixed partners different jobs. Strategy should not pretend to be risk. Finance should not pretend to be customer truth. A contrarian should be allowed to say the move is wrong.

Three design decisions now matter most to me:

1. Disagreement is a feature. The report should show where the partners split, not average them into agreeable mush.

2. Missing context must stay visible. If the Board cannot answer honestly, it asks a bounded follow-up or shows conditional paths instead of inventing certainty.

3. The meeting has to end in a record. The useful output is the decision, reasons, forecast forks, odds, and evidence that would change the call—not another chat transcript.

The web version is ready to test now: https://sudhi.app/workspace

You can create an account, bring a question or a decision you want challenged, add the context you choose, hear the Board's different views, and leave with a decision record. The mobile app is coming next. I am validating the real Android and iOS experience now rather than calling it released early.

The question I am still wrestling with is where candor becomes noise. Would you rather see all ten opinions every time, or start with the strongest disagreement and open the rest only when needed?