{{first_name | Leader}}, welcome back.
Tanmai Gopal, CEO of PromptQL, posted on X that his company had "raised $136M to kill Slack." He shared a product video, challenged Sam Altman to try it, and even promised to donate $100K to charity, plus have his mom make an apology video, if OpenAI tried it and hated it.
The post racked up 2.8 million views, so it's no surprise people noticed. But getting attention isn't the same as being important. Here’s what happened.
The Launch
PromptQL is an AI-first workspace. It looks a lot like Slack, with channels, threads, and search, but what happens behind the scenes is very different.
In Slack, conversations are just messages. In PromptQL, every conversation becomes a useful memory. If someone explains why a dashboard broke or shares a workaround, the AI can remember and use that information later, instead of letting it disappear into another thread that everyone has to dig up or explain again a few weeks later.
This also wasn't a new direction for Hasura. The company was talking about the problem of AI assistants that people don't trust back in January 2025, about a year and a half before this launch.
Why this launch video actually worked
Set aside whether PromptQL replaces Slack for a second. As a launch, this thing did real work. It pulled in 2.8 million views on a Tuesday night, and most product launches never get that kind of attention.
It picked a fight with a specific person. Saying you're better than legacy chat tools is easy to ignore. Telling Sam Altman to try your product creates a story people would want to watch.
The challenge was low risk for PromptQL and high cost to ignore. Offering to donate $100,000 to a charity of Altman's choice if he tried the product and didn't love it meant the company had little to lose if he stayed silent. At the same time, it became an unusual thing to publicly ignore, which helped the launch spread.
The claim of an apology video from the founder's mother made the launch memorable. People rarely share screenshots of product claims. They do share weird, personal details that make a launch feel different and give everyone something to react to.
The lesson is to make one bold, specific claim, attach real stakes to it, and include one memorable human detail, and that's what people shared.
Why this matters right now
Every big company is adding AI to its chat apps. Slack has AI summaries. Microsoft has Copilot in Teams. Everyone has their own version. PromptQL is betting that this approach doesn't go far enough. Slack was built to help people send messages, not to give AI long-term memory.
People can figure out why a launch was delayed by asking coworkers and filling in the gaps. AI can't. If the context isn't there, it has to guess. And it usually sounds very confident, even when it's wrong.
So how is it different?
When you ask PromptQL a question, it doesn't just summarize old chat threads. It pulls the actual data from tools like Snowflake or Salesforce, then shows you exactly where the answer came from.
If someone corrects the answer, it remembers that for the next person who asks. It also learns that words like "churn" can mean different things to different teams, so people don't have to explain the same thing over and over again.
Anyone can build a good-looking demo. What's more interesting is that PromptQL actually stopped using Slack for its own team. It wasn't just a short test. They made the switch for good, and CEO Tanmai Gopal has said the same thing in multiple posts, not just during the launch.
They also have customers like Instacart, McDonald's, and Cisco. Companies at that scale don't adopt new software without carefully checking its security and reliability. The actual use case makes it much more credible than a viral post on X.
Where I push back
Saying "we killed Slack" makes for a great headline, but it's hardly true.
What actually happened is that a company with around 70 employees decided Slack wasn't built for the way they wanted to work with AI. So they built something that worked better for them. That's an interesting case study, but it's not proof that Slack is outdated for the thousands of companies that still use it every day.
The challenge to Sam Altman is similar. It grabs attention, but it doesn't prove anything. It mixes up two different ideas: that there's a real problem with how today's AI collaboration tools work, and that PromptQL has already built the product that will replace Slack.
One more thing: the $136 million figure closely matches the total amount of funding previously raised by Hasura, the company PromptQL came out of. It's not clear whether this is all new funding, earlier funding that's being counted again, or a mix of both. That should be checked before calling it a fresh funding round.
The takeaway
Forget whether PromptQL replaces Slack. The bigger idea is that workplace software might be split into two different roles: one for helping people communicate, and another for giving AI the memory it needs to do useful work.
Slack was built for the first job, and it does that well. What no one knows yet is whether you need a completely new kind of tool for the second job. PromptQL is betting that you do.
Other launches worth a look this week
PromptQL wasn't the only one making noise this week. Take a quick look at these 3 other launches.
Meta launched Muse Spark 1.1, its first paid AI model, undercutting OpenAI and Anthropic on API pricing while chasing them on agentic coding and computer-use benchmarks.
OpenAI launched ChatGPT Work, a GPT-5.6-powered agent that turns a goal into finished slides, sheets, docs, and websites, OpenAI's answer to Anthropic's Claude Cowork.
Prime Intellect raised $130M to build what it's calling the "open superintelligence stack," full-stack infra for training, deploying, and continuously improving your own AI models instead of renting access from a closed lab.
That’s all for today. Let us know if you’d want us to cover any of these launches in detail.
PS. Did you see our AI Tools Directory? It’s a curated collection of 250+ useful AI tools for your profession, business domain, or specific use cases.
Stay curious, {{first_name | leaders}}