Why I’m Joining TextQL

Brian Bickell - Head of Partnerships

I’ve known Ethan and Mark for a few years now, after meeting them how most partnerships folks meet people, by hoping to use them to sell more software and retire partner sourced quota. You might think your partnerships people are nice, but this is how they are wired.

A productive partnership with TextQL never really materialized at my prior gig. Probably, because Ethan wasn’t convinced that an independent, universal semantic layer was worth the time and effort to integrate with. That could be because the distribution of universal semantic layers never hit a tipping point where they were the default state for customers, the way cloud data warehouses have become. If customers don’t ask you to integrate with something, it typically doesn’t get done. Totally fair.

Despite being useless to my partner sourced revenue goals, I kept up with Ethan and hung out with him on the conference circuit, always enjoying his unique combination of high energy, challenging conventions and intellectual honesty. We stayed in touch by text and I followed the progress of the company, always happy to help with advice, or introductions when asked.

Look, I know I told you above that partnerships people are interested in their own goals, but it’s also a fact that the world is small and being a good person pays dividends. Cooperation is a winning strategy.

At some point I’d helped open enough doors that Ethan asked how he could hire me. I didn’t really take it seriously for months because I am not easily swayed. Check my LinkedIn and you’ll find some long stints. What convinced me to take a second look was a combination of the growth metrics of Ana, the TextQL AI agent for data science, as well as the attention that the team’s recent launches were getting.

I got hands on with Ana, asking my own questions and saw something I didn’t think was possible. An AI agent performing deep research level analytics by calling tools and iterating the way a human analyst would, without substantial upfront configuration. Ask Ana a question, and it parses your question for meaning, and intent and then begins writing SQL to introspect your databases to figure out what it’s got access to. It decomposes the question and tries multiple approaches the way I would have when I was a data analyst. It notes findings and updates its understanding of the world along the way. It also happens to do this over some rather gnarly schemas with complex joins and thousands of tables.

After Ana has a handle on the problem it emits useful assets along the way, including record sets, visualizations, reports, and even up to well formatted presentations of findings. Because Ana is good at calling tools, more are added all the time. Ana also has access to a python sandbox and is very good at writing API calls to your other systems.

Ana feels like something different in that it looks like Claude, or ChatGPT, but it can answer real questions over real data, of real complexity within minutes of configuration. Belief that heavy upfront configuration and “getting our data ready” is necessary to deliver AI analytics is killing many initiatives before they are attempted.

The momentum, the reaction from customers, and how it feels to use Ana convinced me. I came aboard TextQL as the Head of Partnerships earlier this week.

In my first week at TextQL the most common answer to my questions about the state of the product, the business, or the users is simply “dude just ask Ana”. It took me just a few hours to internalize that. Ana, connected to our production system data, has proven to be an incredible treasure trove of partner insights.

Immediately I could figure out which consulting firms were already using the product. How often? How much usage? Who is logging in? Did they connect their own data, or are they using our sample data sets?

I wanted to learn about what our largest customers are using the product for so that I could start to generalize stories to tell consulting partners I might recruit. Easy, just ask Ana to characterize all of our production chats by use case. Now I know what our major customers are using TextQL for. I didn’t have to read stale data from a CRM. I didn’t have to go watch the call recordings. I went right to the production data. I never got directed to an out of date visualization with broken assumptions.

I also didn’t have to ask someone here “hey where can I find out more about our customer use cases?”. If you’ve ever worked in early stage startups, you know that a lot of times you’re asking for assets that don’t actually exist. They are stories in the collective, tribal knowledge of the organization. They are also not evenly distributed.

It’s day three for me here and I’m pleased to discover how much use I’m getting out of our own product. If you’ve worked in technology you probably know that this isn’t always the case. Many folks build whole careers around products that they barely use themselves. It’s hard to be excited about that. It’s early days, but I’m excited about what we’re building here.

I’m also excited to get to work bringing Ana to our partners and seeing what they think of it. I can imagine a new generation of system integrators, suddenly able to tackle much more complex business analysis and punch far above their weight than before. As a guy with a soft spot for the small consulting firm, that’s a future I want to be a part of.

If you’re curious about what we’re doing here and think that you or your customers would be interested, please reach out to me and let’s chat. I’d love to show you why I think what we’re building is cool.