Jordan built Mate, a company phone book modeled on Amazon's phone tool, in a weekend with Bolt, Cursor and Supabase. Why remote teams need one.
A company can build its own internal phone book app in a weekend with AI coding tools, and for a remote team it becomes a real culture tool. On Episode 5 of Built This Week, Jordan Metzner showed Mate, the company directory he built for Ryz Labs with Bolt, Cursor and Supabase. Sam Nadler says any leader, technical or not, could build something similar.
First, the name. Jordan explains that mate is an Argentinian tea drink that is common in Argentina, Uruguay and Brazil and is usually shared among friends. Spelled in English, mate is also Australian for friend.
The idea came from Amazon, where Jordan used to work. Amazon had an internal phone book, which people called the phone tool. It showed who else was on your team, who your manager was and who your manager's manager was, all the way up to Jeff Bezos and back down again. It also let people join communities, list interests and earn badges. Jordan says it became a very important tool inside the company and in creating company culture.
He had wanted to build one for a long time. At first the company was too small, and later it was never the top priority. AI changed the math. In his words, it allowed him to develop software so much faster that he built the whole app over a few days on a weekend. He says it is not an overly complicated app, and he thinks most companies could build one for themselves.
Watch Jordan explain why we built Mate in Episode 5 (starts at 1:53) · Episode 5 notes and transcript
Sam says we think culture is important, and the team is fully remote:
We also have a fully remote team. So we're constantly looking for ways to create, moments of connection, to create things to bond over, and that can be really limiting in the remote environment.
— Sam Nadler, Episode 5
At a previous job he had a similar tool called the family tree, and it was one of the tools he used most to learn about teammates or people he met in meetings. Mate does something similar. People share pictures, the app shows when they started so you can see tenure, and it lists interests, so two people can bond over a favorite football team or music group.
Mate opens on your own profile. Sam walked through Jordan's: a profile photo, then badges for the company offsites he attended. Below that are interests, and other teammates can join each one. Jordan's included Bitcoin, yoga and surfing, and Sam was surprised how many people had joined surfing. There are also clubs. One dog group has a lot of people in it, and quite a few teammates have joined Boca Juniors, the famous soccer team.
At the bottom of each profile are personal photos. Sam shared pictures of his two daughters and his puppy, and Jordan's page shows his dogs and his wife. Sam is clear that this is not a groundbreaking product, just an easy tool that anyone can create without being very technical. He says it had been live for a couple of months at the time and people use it all the time.
Jordan thinks it is especially helpful for people who are just starting and don't know many coworkers yet. They can find people with similar interests and bond with them. The app has badges, interests, clubs and teams, plus search, so you can look up anyone you work with by name.
Sam adds that for new teammates it is a way to get familiar with different members of the team and with the company's cultural ethos. Jordan expects to keep improving it over time to make it a better and better experience. He adds that the first thing you notice is that you start seeing people's dog and cat photos and learn more about your coworkers than you would on a normal day.
Jordan built the first version in Bolt (bolt.new), which he describes as an application that lets you build front ends. He used it to scaffold the start of the app and lay out the left-hand sidebar. Bolt integrates with Supabase directly, so he could then take the app out of Bolt and finish it in Cursor.
Supabase is the database of record, and Jordan also uses it for search, authentication and a number of small features from the Supabase ecosystem. He admits the app was built a while ago and probably deserves a fresh coat of paint. We have a separate page on the Mate build, and Bolt also appears in our Bolt vs Claude Code comparison.
I think it's a tool that any leader, technical or nontechnical, can probably spin up in a weekend and have a fun kind of cultural tool for their team to enjoy.
— Sam Nadler, Episode 5
In the news segment, Jordan called Google's newly shared numbers bonkers. Google said over 50,000,000 people are using AI-powered meeting notes inside Google Meet. He still sees OpenAI as the leader, but thinks the numbers show two things: Google is catching up, and there is massive consumer demand to use AI models every day.
So it just shows the power of Google's network, how many users they have, and, once they put AI in front of users, how fast it gets adopted.
— Jordan Metzner, Episode 5
He pointed out that Google went from what he would call last place in having a successful LLM to the number two or three spot.
Sam thinks so. On Episode 1 we covered Granola, a meeting notes tool, and now Google has layered the same idea into Meet with 50,000,000 users. Sam says both of us are pretty AI forward, and he is not sure he has even used the Meet feature yet.
His point is that Google has more places to put AI in front of people than OpenAI does: inside Google Docs, Gmail and search, AI is at people's fingertips, while OpenAI users have to be more intentional. He thinks that will help Google close the gap with Gemini. Jordan made the same case for Meta: it has so many eyeballs on its products that once it integrates AI into them, he expects a huge amount of adoption.
The episode includes Sam's walkthrough of real Mate profiles, interests and clubs, and the full Google discussion. Watch Episode 5 of Built This Week, where the full transcript is also on the page.
