How the Smartest Companies Use AI | Ankur Goyal, Braintrust
The Peel with Turner Novak
Turner Novak
4.6 • 11 Ratings
🗓️ 7 June 2024
⏱️ 87 minutes
🧾️ Download transcript
Summary
Get Attio, the next generation of CRM: https://bit.ly/AttioThePeel
Ankur Goyal is the Founder and CEO of Braintrust, the end to end developer platform for building the world's best AI products. Their customers include companies like Instacart, Zapier, Notion, Airtable, Replit, and more.
We hit on the importance of LLM evals, advice for building AI products, why the best companies have two AI product roadmaps, and his non-conventional advice for founders.
Timestamps:
(00:00) Intro (04:04) Why everyone’s now an AI company (06:03) Reasons LLM evals are so important (08:10) Typescript becoming the language of AI (09:19) Replacing vibe checks with Braintrust (10:37) Making OpenAI’s protocols the standard (11:27) Why the best companies have two AI roadmaps (13:06) Building your product so each LLM release makes it better (14:54) Predicting AGI is impossible (15:54) Why people who work with LLMs aren’t worried about AI safety (16:52) The best developers are all-in on co-pilots (18:11) How AI is changing software development (21:09) Combining IDE, CIDC, and observability in one product (27:18) Are models more like CPU’s or relational databases? (30:14) How to pick an LLM (33:00) Advice for staying on top of new AI developments (34:30) Why tool calling is so important (38:02) Advice for young software engineers (40:25) Learning to code doing linear algebra homework (42:36) Lack of purpose interning in big tech (44:07) Working at MemSQL learning to be a founder (47:52) How to get a job at a startup (50:43) Building his first startups product on an international flight (52:39) Three lessons from his first failed startup (54:46) Don’t delegate what you’re good at (55:46) Why you should be careful listening to VCs advice (57:34) Tactics for successful delegation (59:36) Why Ankur doesn’t do any meetings (01:02:42) The importance of self-service in unlocking certain customer segments (01:05:14) How Braintrust got started (01:07:45) Advice on picking your target customers (01:10:35) How Braintrust hires with work trials (01:15:21) Balancing security with a modern UI (01:17:49) Why it’s hard to sell non-AI products right now (01:19:21) Advice for selling to large enterprises (01:23:10) Ankur’s favorite AI products
Referenced:
https://www.braintrustdata.com/
SICP Book
PDF: https://web.mit.edu/6.001/6.037/sicp.pdf
Hardcopy: https://www.amazon.com/Structure-Interpretation-Computer-Programs-Engineering/dp/0262510871
Linear’s guide to work trials: https://linear.app/blog/why-and-how-we-do-work-trials-at-linear
Where to find Ankur:
Twitter: https://twitter.com/ankrgyl
LinkedIn: https://www.linkedin.com/in/ankrgyl/
Where to find Turner:
Twitter: https://twitter.com/TurnerNovak
LinkedIn: https://www.linkedin.com/in/turnernovak/
Newsletter: https://www.thespl.it/
Transcript
Click on a timestamp to play from that location
| 0:00.0 | What the smartest companies are doing right now is they basically have two product roadmaps. |
| 0:05.9 | One roadmap assumes that AI is going to keep getting better, |
| 0:10.9 | but it's not going to be what people scientifically think of as AGI in the next five or 10 years. |
| 0:17.5 | And then they have another roadmap, which assumes that all these crazy things happen. |
| 0:21.1 | You want to structurally build a product so that if Open AI or Google Anthropic, et cetera, |
| 0:28.2 | if they release a model that is twice as good as the model today, your product gets better. |
| 0:34.0 | Most of our customers are forecasting at least 50% of their engineering projects over |
| 0:38.9 | the next year or two involving AI and probably 100% in the next five years. What we're seeing |
| 0:44.6 | is that the paradigm around building software is totally different. It no longer revolves around |
| 0:50.2 | code sitting in an IDE. It revolves around data that you get from your users and prompts that, |
| 0:58.4 | you know, not just engineers, but product managers, users, you know, other people are able to |
| 1:02.7 | contribute to control and affect how the software actually works. I hate meetings. So at BrainTrust, |
| 1:10.1 | we have one company meeting per week. |
| 1:12.1 | Okay. Wow. For a week. I thought you might say day. No, no, no. It's one per week. And the longest it's |
| 1:17.6 | ever taken is 15 minutes. Folks at larger companies will chuckle and they should. But we give people an |
| 1:23.3 | insane amount of autonomy. And, you know, I'm actually the type of person who believes it's better |
| 1:28.6 | to give people a lot of autonomy and for them to mess up occasionally or be slow or unproductive |
| 1:34.5 | for a week while they're wrapping their head around a problem than to micromanage people. |
| 1:41.7 | Welcome to the Peel, where we explore the world's greatest startup stories. |
| 1:45.9 | I'm your host, Turner Novak, founder of the native capital, a venture capital firm that also |
| 1:50.0 | hates meetings. |
| 1:51.1 | Today, I talked to Anker Goyle, founder and CEO of Braintrust, the end-to-end developer |
... |
Please login to see the full transcript.
Disclaimer: The podcast and artwork embedded on this page are from Turner Novak, and are the property of its owner and not affiliated with or endorsed by Tapesearch.
Generated transcripts are the property of Turner Novak and are distributed freely under the Fair Use doctrine. Transcripts generated by Tapesearch are not guaranteed to be accurate.
Copyright © Tapesearch 2026.