Project: Ship Agent v1 with a build log

Intermediate⏱ 35 min🛠 Practical project
🔒Join free to watch this video lessonFree account · takes 30 seconds · no card neededJoin free & watchAlready a member? Log in
Deploy, test with at least ten calls, measure latency and document everything you changed.

In this lesson you’ll learn to

  • 10+ logged calls
  • Latency measured
  • Build log written

Your deliverable

• A working phone number that reaches your agent (inbound) — and, optionally, an outbound test call flow to your own phone.

• Test log of at least 10 calls with scenario, result, issue and fix.

• Latency measurement. Stopwatch or use platform analytics: average pause between caller finishing and agent starting, over at least 10 turns. Report the average and the worst case.

• Two recordings (with consent): one good call and one failure you learned from.

• One-page build log: stack chosen and why, settings changed, top five problems found, how you fixed each, what you would do next.

Quality bar

• Average response gap is measured, not guessed.

• The agent introduces itself as an AI assistant.

• Out-of-scope questions g

🔓 Unlock the full lesson — free

Create a free account to read the step-by-step guide, see the worked example, take the quiz, download templates and track your progress.

💬 Questions about this lesson?
Ask the community — practitioners and the BFOUNT team answer.
Ask a question
Was this lesson helpful?