Free tool
What will your audience ask?
Paste the topic or the abstract of your talk, and get the questions people are likely to ask, so you can prepare the answers before you go on.
Worked example
A sample abstract and 8 likely questions
We wrote this example by hand to show the format.
Cutting API latency in half: what we changed in our gateway
This year the p95 latency of our public API fell from 480 ms to 210 ms. This talk walks through the three changes that did most of the work: a connection pool in front of Postgres, a rate limit per tenant instead of one global limit, and auth checks at the edge. For platform teams who run a multi-tenant API.
- Numbers Is the drop from 480 ms to 210 ms for every tenant, or mostly for the largest ones?
- Design Why a limit per tenant and not one global limit, and what happens to a tenant that sends a burst?
- Design How do you size the connection pool, and what happens when it runs out?
- Risk With auth at the edge, how fast does a revoked token stop working?
- Numbers Which of the three changes gave the biggest win on its own?
- Cost Did the changes raise your infrastructure bill?
- Rollout How did you ship this without downtime, and how would you roll it back?
- Hindsight What would you change if you started again today?
In the extension, the answers come too
Vagmi has this feature today. In settings, "Prepare for a talk" writes the likely questions from your topic or from your material, and answers each one from your sources, with citations.
During the call, a question that matches one of them gets the full prepared answer right away. The Free plan includes 1 prepared pack, one time.