Get live statistics and analysis of Mike Ritchie's profile on X / Twitter

I've taken inspiration from products I use (including @linear), but this is fucked up.

ya know, @duckdb is open source, you could help instead of complain

We (@definiteapp) have a PR open to add predicate pushdown to the DuckDB Iceberg extension. This is a 0 to 1 change for many production @ApacheIceberg lakes. @duckdb currently downloads every parquet file for every table referenced. This is... not ideal. Most queries would just bomb. With predicate pushdown, we see: * simple queries: from several seconds --> subsecond * complicated queries: From OOM errors --> query actually executes

Took DuckLake for a spin this weekend and I was not disappointed. @duckdb nailed the experience we've been painfully stitching together at Definite with Iceberg + DuckDB.

We know @duckdb is high-quality, performant software. But I didn't appreciate how well the people building it are organized. Here's a section of the 1.0.0 release announcement: DuckDB Labs: The core contributors. Instead of developing features that will be behind a paywall, they provide support and consulting. Awesome. DuckDB Foundation: A non-profit that ensures DuckDB remains MIT licensed. Perfect.

We switched our CRM at @definiteapp to @attio last week. It took me from: dislike -> like I dreaded logging into our old CRM. I enjoy opening up Attio. The result? I spend more time selling. As a technical founder, I always want to work on product. Having good tech to work on sales helps shift the balance in the right direction. 𝐇𝐞𝐫𝐞'𝐬 𝐰𝐡𝐚𝐭 𝐈 𝐥𝐢𝐤𝐞: 𝐓𝐡𝐞 𝐭𝐚𝐛𝐥𝐞𝐬 / 𝐯𝐢𝐞𝐰𝐬 𝐚𝐫𝐞 𝐩𝐞𝐫𝐟𝐞𝐜𝐭𝐥𝐲 𝐝𝐨𝐧𝐞. Adding columns, filtering, and sorting all work intuitively. You can edit records without opening them (e.g. you can edit the value for 3 records from the table view, without opening each one) and everything syncs automatically. (I could be bias here, because the tables work exactly like they do in Definite... I think we may have both been inspired by Notion 😂 ) 𝐓𝐡𝐞 𝐤𝐞𝐲 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧𝐬 𝐚𝐫𝐞 𝐛𝐮𝐢𝐥𝐭-𝐢𝐧 𝐚𝐧𝐝 𝐭𝐡𝐨𝐮𝐠𝐡𝐭𝐟𝐮𝐥𝐥𝐲 𝐢𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐞𝐝. You connect your email account and Attio (without relying on bcc) automatically pulls in leads and calculates a "Connection Strength". They also automatically pull in stuff you'd normally need to get from Clearbit (e.g. amount raised, # twitter followers, etc.) 𝐒𝐩𝐞𝐞𝐝 𝐚𝐧𝐝 𝐔𝐗. The app loads fast and all user interactions are snappy. I have no affiliation with them, I just respect craftsmanship like this.

Claude artifacts would be 10x more interesting if they added @supabase. My wife had an idea for a wordle like, two player dice game. Had it running in Claude in a few minutes, but it broke as soon as I tried to make it real. @kiwicopple have you talked to them?

our post on @duckdb + @ApacheArrow made the front page

We just open-sourced our @meltanodata target for Ducklake (from @duckdb). * Type conversion is automatic (timestamps stay timestamps). * Append or merge: choose at runtime. * Storage is portable (S3, GCS, or local). * Works with Postgres, MySQL, SQLite, or DuckDB catalogs. * Timestamp and categorical partitions are built in. We're already running it in production here at @definiteapp.

Excited to join @jrdntgn (MotherDuck), Jake Thomas (Okta), and @mehd_io (MotherDuck) for a panel on scaling DuckDB—from single-node setups to distributed architectures. We’ll cover: • Should you care about DeepSeek’s smallpond? (sneak peek: “maybe”) • When to scale up vs. scale out • Real-world experiences with @motherduck, self-hosted DuckDB, and large-scale workloads 📅 Tuesday, March 11 ⏰ 8:30am PT / 4:30pm CET 🔗 Register here: lu.ma/5946jam3 If you're working with @duckdb, don’t miss this!

Another post by @stevowang on querying @ApacheIceberg tables with @duckdb in a cloud setting using @awscloud. Full post here: definite.app/blog/cloud-ice….

Usage of @duckdb is up 230% YoY in the @StackOverflow survey

It shouldn’t take 6 months and 5 vendors to make sense of your data. @definiteapp raised $10M to give you an AI-native data platform in an afternoon. Why does it normally take so long? Because the modern data stack is split across three heavyweight products: 1. Data Warehouse: a place to store all your data 2. ETL: pipelines to get data into your warehouse 3. BI: a place to build reports & dashboards Each of these comes from a different vendor, each with its own price tag and complexity. On top of that, you need data engineers and analysts to stitch them together and keep the system running. Let’s look at one of our customers, a b2b SaaS that raised $20M. Their current tech stack looks like many of our customers: * Postgres for their application database * Hubspot for CRM * Stripe for payments * A slew of other sources (customer success platform, Google Sheets, and event data) They wanted the same thing every company wants: a single source of truth for how the business is performing. Traditionally, this is done with dashboards. Their plan was to hire a few data people, buy a bunch of tools (Snowflake, Tableau and Fivetran) and hope for the best. Six months and a lot of pain later, maybe they’d have a few dashboards. Fortunately, they tried Definite one morning and by that afternoon their key metrics were live. How do we do it? Definite ships a full data stack (warehouse + pipelines + metrics & reports) in a single app. We’re built on the best open source infrastructure in the market with an AI agent that’s eager to make sense of your data. Instead of hiring, buying, stitching, and waiting; you have the answers you need in minutes running on a data stack that scales. And it's just a chat away. A huge thank you to our partners at @costanoavc (@JohnCowgill) and @AcrewCapital (@asadkkhaliq), to our incredible team, and to our early design partners who have helped shape the product. And we’re hiring. If you’re a high-agency builder who wants to help reinvent analytics for thousands of operators, let’s talk —> mike@definite.app

⚫ Definite 2.0 This is the biggest release since we started Definite and I’m insanely proud of it. We’ve completely redesigned the product with meticulous attention to every detail. We’ve added scalable storage to @duckdb using @ApacheIceberg. And our AI Assistant (”Fi”) is now ready to help you get shit done. It was painstaking and took months to get right, but the result is worth every minute. Incredibly proud of the team here for pulling this off.

I've taken inspiration from products I use (including @linear), but this is fucked up.

We switched our CRM at @definiteapp to @attio last week. It took me from: dislike -> like I dreaded logging into our old CRM. I enjoy opening up Attio. The result? I spend more time selling. As a technical founder, I always want to work on product. Having good tech to work on sales helps shift the balance in the right direction. 𝐇𝐞𝐫𝐞'𝐬 𝐰𝐡𝐚𝐭 𝐈 𝐥𝐢𝐤𝐞: 𝐓𝐡𝐞 𝐭𝐚𝐛𝐥𝐞𝐬 / 𝐯𝐢𝐞𝐰𝐬 𝐚𝐫𝐞 𝐩𝐞𝐫𝐟𝐞𝐜𝐭𝐥𝐲 𝐝𝐨𝐧𝐞. Adding columns, filtering, and sorting all work intuitively. You can edit records without opening them (e.g. you can edit the value for 3 records from the table view, without opening each one) and everything syncs automatically. (I could be bias here, because the tables work exactly like they do in Definite... I think we may have both been inspired by Notion 😂 ) 𝐓𝐡𝐞 𝐤𝐞𝐲 𝐢𝐧𝐭𝐞𝐠𝐫𝐚𝐭𝐢𝐨𝐧𝐬 𝐚𝐫𝐞 𝐛𝐮𝐢𝐥𝐭-𝐢𝐧 𝐚𝐧𝐝 𝐭𝐡𝐨𝐮𝐠𝐡𝐭𝐟𝐮𝐥𝐥𝐲 𝐢𝐦𝐩𝐥𝐞𝐦𝐞𝐧𝐭𝐞𝐝. You connect your email account and Attio (without relying on bcc) automatically pulls in leads and calculates a "Connection Strength". They also automatically pull in stuff you'd normally need to get from Clearbit (e.g. amount raised, # twitter followers, etc.) 𝐒𝐩𝐞𝐞𝐝 𝐚𝐧𝐝 𝐔𝐗. The app loads fast and all user interactions are snappy. I have no affiliation with them, I just respect craftsmanship like this.

ya know, @duckdb is open source, you could help instead of complain

⚫ Definite 2.0 This is the biggest release since we started Definite and I’m insanely proud of it. We’ve completely redesigned the product with meticulous attention to every detail. We’ve added scalable storage to @duckdb using @ApacheIceberg. And our AI Assistant (”Fi”) is now ready to help you get shit done. It was painstaking and took months to get right, but the result is worth every minute. Incredibly proud of the team here for pulling this off.

It shouldn’t take 6 months and 5 vendors to make sense of your data. @definiteapp raised $10M to give you an AI-native data platform in an afternoon. Why does it normally take so long? Because the modern data stack is split across three heavyweight products: 1. Data Warehouse: a place to store all your data 2. ETL: pipelines to get data into your warehouse 3. BI: a place to build reports & dashboards Each of these comes from a different vendor, each with its own price tag and complexity. On top of that, you need data engineers and analysts to stitch them together and keep the system running. Let’s look at one of our customers, a b2b SaaS that raised $20M. Their current tech stack looks like many of our customers: * Postgres for their application database * Hubspot for CRM * Stripe for payments * A slew of other sources (customer success platform, Google Sheets, and event data) They wanted the same thing every company wants: a single source of truth for how the business is performing. Traditionally, this is done with dashboards. Their plan was to hire a few data people, buy a bunch of tools (Snowflake, Tableau and Fivetran) and hope for the best. Six months and a lot of pain later, maybe they’d have a few dashboards. Fortunately, they tried Definite one morning and by that afternoon their key metrics were live. How do we do it? Definite ships a full data stack (warehouse + pipelines + metrics & reports) in a single app. We’re built on the best open source infrastructure in the market with an AI agent that’s eager to make sense of your data. Instead of hiring, buying, stitching, and waiting; you have the answers you need in minutes running on a data stack that scales. And it's just a chat away. A huge thank you to our partners at @costanoavc (@JohnCowgill) and @AcrewCapital (@asadkkhaliq), to our incredible team, and to our early design partners who have helped shape the product. And we’re hiring. If you’re a high-agency builder who wants to help reinvent analytics for thousands of operators, let’s talk —> mike@definite.app

Usage of @duckdb is up 230% YoY in the @StackOverflow survey

Claude artifacts would be 10x more interesting if they added @supabase. My wife had an idea for a wordle like, two player dice game. Had it running in Claude in a few minutes, but it broke as soon as I tried to make it real. @kiwicopple have you talked to them?

We (@definiteapp) have a PR open to add predicate pushdown to the DuckDB Iceberg extension. This is a 0 to 1 change for many production @ApacheIceberg lakes. @duckdb currently downloads every parquet file for every table referenced. This is... not ideal. Most queries would just bomb. With predicate pushdown, we see: * simple queries: from several seconds --> subsecond * complicated queries: From OOM errors --> query actually executes

two steps: * connect to your sources (hubspot, stripe, ads, etc.) * tell @definiteapp what you want to see

Another post by @stevowang on querying @ApacheIceberg tables with @duckdb in a cloud setting using @awscloud. Full post here: definite.app/blog/cloud-ice….

Excited to join @jrdntgn (MotherDuck), Jake Thomas (Okta), and @mehd_io (MotherDuck) for a panel on scaling DuckDB—from single-node setups to distributed architectures. We’ll cover: • Should you care about DeepSeek’s smallpond? (sneak peek: “maybe”) • When to scale up vs. scale out • Real-world experiences with @motherduck, self-hosted DuckDB, and large-scale workloads 📅 Tuesday, March 11 ⏰ 8:30am PT / 4:30pm CET 🔗 Register here: lu.ma/5946jam3 If you're working with @duckdb, don’t miss this!

Why do people hate "x for y" taglines? e.g. @cursor for doctors

Good read here on how and why @NotionHQ built a data lake. tldr: they saved over $1M by moving off Fivetran + Snowflake. They also set up CDC (change data capture), so ingestion times dropped from days to hours. If you read this and think "wow, that sounds great, but we'd never build this ourselves", check out @definiteapp (spoiler: we built it for you).