16 Sep
28 Oct
24 Nov
15 Dec
16 Jan
9 Feb
18 Mar
22 Apr
20 May
17 Jun
22 Jul
19 Aug
16 Sep
21 Oct
18 Nov
16 Dec
13 Jan
10 Feb
9 Mar
13 Apr
11 May
8 Jun
13 Jul
10 Aug
14 Sep
12 Oct
9 Nov
14 Dec
11 Jan
8 Feb
8 Mar
12 Apr
10 May
14 Jun
12 Jul
9 Aug
13 Sep
11 Oct
8 Nov
13 Dec
10 Jan
14 Feb
14 Mar
11 Apr
9 May
13 Jun
11 Jul
8 Aug
12 Sep
10 Oct
14 Nov
12 Dec
9 Jan
13 Feb
13 Mar
10 Apr
8 May
12 Jun
10 Jul
14 Aug
11 Sep
9 Oct
13 Nov
11 Dec
8 Jan
12 Feb
11 Mar
8 Apr
13 May
10 Jun
8 Jul
12 Aug
9 Sep
14 Oct
10 Nov
9 Dec
13 Jan
10 Feb
10 Mar
14 Apr
12 May
9 Jun
14 Jul
11 Aug
10 Sep
13 Oct
10 Nov
15 Dec
12 Jan
9 Feb
13 Apr
11 May
8 Jun
13 Jul
10 Aug
14 Sep
12 Oct
9 Nov
14 Dec
13 Jan
8 Feb
12 Apr
10 May
14 Jun
12 Jul
9 Aug
13 Sep
11 Oct
8 Nov
13 Dec
10 Jan
14 Feb
13 Mar
8 May
12 Jun
10 Jul
14 Aug
11 Sep
9 Oct
21 Nov
11 Dec
8 Jan
12 Feb
12 Mar
9 Apr
14 May
11 Jun
9 Jul
13 Aug
10 Sep
8 Oct
12 Nov
9 Dec
14 Jan
11 Feb
11 Mar
8 Apr
12 May
10 Jun
8 Jul
12 Aug
8 Sep
14 Oct
11 Nov
Meetup #143
Event Details
Speakers
State of Voice AI with JS 
Voice agents feel magical until someone talks over them. This talk starts with a working agent built on LiveKit, then takes it apart: how the pieces fit together, and where the whole thing tends to fall over. We'll go through latency — where the milliseconds actually go across STT, LLM, and TTS, and which ones you can buy back; interruptions — barge-in, turn detection, and what to do with a half-spoken sentence; stability — how to move fast with agent development; and guardrails — keeping an agent on topic when the input is unscripted human speech. A live demo, an architecture walkthrough, and the mistakes you don't need to repeat.
Testing the Test Writer: Building and Evaluating an Agentic Playwright Skill
AI agents can generate Playwright code, but producing a reliable E2E test requires much more than writing selectors and assertions. The agent must understand the repository, reuse existing fixtures and page objects, configure the test environment, and prove that the result actually works.
In this talk, I'll show how we built a Playwright test-generation skill at Dropbox that takes a test scenario from a natural-language description to a validated test and how deterministic validation gates make the workflow more reliable. I'll then explain how we evaluate the skill on realistic scenarios and continuously monitor it for regressions. You'll leave with a practical approach for building agentic coding workflows that can be tested and improved like any other production system.
JebTo.AI and Other Bullshit: A Case Study of Professional AI Slop Donkeys in the Age of the Fourth Industrial Revolution
The AI revolution created new tools, new industries, and one of the greatest markets for bullshit in modern history. Suddenly everyone is an expert: a few tutorials, a new LinkedIn headline, and a vocabulary of disruption and transformation. Mike has spent years making award-winning films with AI, which makes him either the best or the worst person to guide you through the sloppocalypse. Through JebTo.AI he dissects the economy of hype, slop and confidence without competence, argues that judgment is now the scarce resource, and gives an honest, behind-the-scenes look at what actually works and what just makes noise. Bring your sense of humour.
- 144Meetups
- 293Speakers
- 491Talks