The work is to see
more clearly.

Founding a company means deciding before you have all the information. More activity is easy to measure. Better judgement is harder — and more useful.

A partner, not an applause machine.

A good partner is on your side and still willing to disagree. That means precise questions, clear reasoning, and respect. It does not mean aggression, shame, or assuming every obstacle is an excuse.

Start from what we actually know.

A founder's account, a customer comment, a payment, and repeated usage tell us different things. We keep those distinctions visible. An attached link is a source to inspect, not an automatic verification of truth.

Make the next move small enough to test.

Every useful recommendation should help you decide or act. State the purpose, the relevant trade-off, and what would change the recommendation. Then choose a step with an observable outcome.

Remember the reasoning, not just the task.

The decision you made three weeks ago may have been right for the evidence you had. New evidence can justify changing it. Keep the earlier decision and its reason so learning survives a change of direction.

Fit the company. Respect the founder.

Deep-tech experiments can take months. A profitable solo business may not need investment. Different companies need different clocks and different definitions of progress.

Privacy is part of the product.

Founders need room to think before they are ready to present. A mentor should see the brief the founder chose to share, not a hidden summary of private uncertainty.

Earn the return visit.

A streak isn't the goal. We want the next conversation to be useful because it starts from what happened, recognises what was learned, and makes the next decision clearer.