About Pydantic
Pydantic is the most widely used data validation library for Python, downloaded more than TypeScript and relied on by developers at companies like OpenAI, Google, Meta, Netflix, and Nvidia. Founded in 2022 by Samuel Colvin, the company raised a $12.5M Series A led by Sequoia Capital in 2024, bringing total funding to $17.2M. In addition to the open source library, Pydantic offers Logfire, an observability platform with first class support for AI.. The company is headquartered in London with a team of around 30 people.
Chris Samiullah is Pydantic's VP of Partnerships. A former engineering manager at Babylon Health and technical lead with deep roots in Python and machine learning, Chris oversees sales at Pydantic. His purview includes managing the sales pipeline, understanding customer needs, preparing materials for investors, and coordinating across engineering, marketing, and leadership.
Before Day AI, Pydantic was using a CRM called Attio. Chris found it clunky, overly manual, and not AI-native. When Bogomil Balkansky from Sequoia suggested they check out Day AI, Chris decided to give it a shot. Within two months, he was so impressed that he volunteered to do a testimonial.
What was the specific reason Attio was not working for you, and what inspired you to look for something new?
It just felt clunky, and a lot of things I had to do manually. It didn't feel very AI-native. It didn't do a lot of stuff proactively. Everything I had to initiate myself. There was no ability to automate specific actions that I might want to do on a recurring basis, and no ability to access things through an agent to integrate with some of my own workflows. It just felt like something from the previous generation.
Walk me through a specific moment where Day AI changed how you work.
I have to prepare a lot of information to help us understand what's going on. Whenever we're doing presentations to potential investors or trying to understand our customers more deeply, it's possible to generate Day AI Pages which often result in really useful insights and wow moments. For example, if I want to drill down into our ICP, I can ask the Day AI Assistant to pull up all the companies who've been through the sales pipeline and analyze them by industry. At the end, I get really clear statistics about which industries convert better than others.
Summarizing and crunching that across months of data gives you these really valuable distinctions. You can say, "Actually, it turns out with this particular industry we have a harder time, and maybe there's something we need to fix, or maybe we should just avoid it and save our energy for something else." This is really useful for improving the sales motion.
Or maybe it's typical sales organization statistics you want to track. What are your PoC conversion rates? You can just ask the Day AI agent to pull together, over the past six months or whatever time frame you want, what sort of conversion rate we've been seeing on the PoC. This is all stuff I can do with a single prompt, and it pulls together a page that previously would have been hours of crunching and manual work. It's enormously powerful and useful.
Are you using Skills, and what do you have set up?
Yeah, my Day AI Assistant goes through all the meetings that are scheduled for the day and tells me what the previous interactions are on those meetings. It gives me a heads up on what the relationship is, if there's any specific outstanding actions to follow up on, what topics are likely to come up. And if it's the first meeting, it gives me an overview of the company. So again, big time saving on research, and it also means I show up at all the relevant meetings well prepared and haven't forgotten some important interaction from a previous meeting or with a colleague. It allows me to be prepared and just go into every meeting feeling more confident.
How would you explain Day AI to a peer at a different company who is using a traditional CRM?
It's like Cursor for CRM is probably how I'd describe it, because I tend to deal with technical people so they understand what I mean by that. Or maybe Claude Code for CRM, although I think the Cursor analogy is more accurate in terms of how the UI behaves. It's just AI-native, and your team at Day AI has thought about not just the user experience, not just the developer experience or the salesperson experience, but also the agent experience, which I think is increasingly very important. It's not just for you, but it's for your team of agents as well.
When you tell people about Day AI, is there anything specific that surprises them?
I think people on my team who I've showed have been consistently surprised by how little is expected to be done in the UI. It's almost all driven from the agent chat prompt, which most of the time is great. Sometimes it's confusing, but once you get your head around the fact that that's how things are supposed to be done, it becomes much more of an intuitive approach.
And this isn't in terms of how I describe it to other companies, but vendors of ours who I work with, who I want to behave more like Day AI, I will point to the CRM and say, "Do something more like this." For example, with our compliance provider, which is an important bit of software for us, I would like to have to do less in the UI. I would like for more things to just happen and get updated automatically. Clicking around a lot in the UI in this day and age feels kind of barbaric. If we can avoid it, let's avoid it. And your team and your product definitely embodies that. So I like to use it as an example.
If you were to try to replace what Day AI does for you, what would you have to do?
We'd have to pull together a whole load of agents to pull in data from email, Slack channels, interactions, call transcript recorders, pull all that information together, store it effectively for searching, add a bunch of agents on top of it, add a bunch of functionality on top of that data, build a whole bunch of UI. We'd have to vibe code a pipeline, build a whole bunch of automations to understand when to update that custom UI. It would be quite a fiddly operation. I mean, we're an incredibly technically capable team, so we could do it, but we don't want to.
What benefits have your other teammates noticed from the product?
Definitely for marketing and DevRel, better understanding of the ICP and the pain points of the customers. For sharing with the team how we're doing in terms of which deals we close more effectively, how long things take to progress through the pipeline. That's always useful not just for the engineers, but for leadership as well, and for investors. So lots of options there.
Competitor analysis has been very useful. We go through transcripts with prospects where we've been compared against competitors and pull together insights about what the perceptions are, whether they're true or not. Perception matters. Pulling together those sorts of comparisons is really easy in Day AI.
There's all those things around understanding the customer, understanding their perceptions, understanding their pain points. You can pull those insights into prioritization, thinking about what you need to work on in terms of your own product and where to focus sales efforts and marketing efforts. It has a ripple effect across the whole company.
What number would you put on the benefit from Day AI?
Hard to say. I don't have an easy way to crunch that without just making up a number for the sake of making up a number. But I know that I'd be very upset if I was told that I had to switch systems, and I would fight that tooth and nail.
What's an example of something you've come to understand as a result of using the product?
Understanding the customer profile, understanding where we do well, which deals convert. Getting a better sense of our processes, where we're efficient, where we're not, where things fall through the cracks, where we're strong. Just quantifying these things. We had a sort of nebulous, intuitive sense of it before, but now it's very easy to quantify. And that's attractive and useful, because once you can quantify it, you can track it. Once you can track it, you can make it better. A good system allows you to improve processes and ultimately improve deals closed and revenue.
It also gives you this sense of confidence. You're not worried that, "Oh, we haven't set up the data model correctly so we're not going to be able to crunch this information if we need it quickly." It's like, no, it's in the AI. We'll probably be able to figure it out. We'll probably be able to search through it relatively easily. That's an intangible that's worth quite a lot.
