@srbhr Thanks! Honestly, we looked around at our own team and despite being really good at connectors thanks to our MCP work, we had no agents running to help us build the business.
We paused development work for a cycle (2 weeks) and everyone built agents and saw all the challenges with making them actually work. And that led su to what Tadata has become!
the "reads the room" part is what caught me. most slack bots just sit there until you @ them, so the fact that this one actually picks up context is a diffrent thing entirely.
question from someone who lives in slack all day...how does it handle channels where the convo is half decisions and half noise? does it need to be tagged or is it listening in the background?
also curious how deep the 'learns your company' part goes after a week vs after a month.
@bencohen When you are DMing with Tadata, you don't need to tag it explictly. When you are in public channels and you want it to act, you do need to tag it. That also means channels that are a mix of noise and decisions are totally fine.
Learning happens from the very first interaction when it learns your company and your role (based on what you tell it). We see it start to provide helpful connecting the dots comments within the first week. But it grows over time of course.
Really like the restraint here. An AI employee that knows the company context, has access to the right tools, and knows when not to speak feels much more useful than another chatbot living in Slack. Congrats on the launch!
the delegation pitch is appealing but what's the failure mode - if Tadata researches a lead and gets something confidently wrong (stale data, wrong company, a hallucinated detail), does it flag its own uncertainty, or does that just show up as a normal Slack message and get treated as ground truth by whoever asked?
Report
@galdayan Great question! If Tadata isn't certain it will ask the user and remember the preferences, when the answer is more complex than a simple slack message it'll create a deliverable like an attachment or a google doc/sheet where it can also detail things like process and level of confidence. like any AI agent, it can still be confidently wrong from time to time, we encourage users to read and validate results, just like you would with your human employees
apideck
Congratulations @tori_seidenstein and the team. I love how the product has evolved from just MCPs to an AI Employee in Slack. ✨
Tadata
@srbhr Thanks! Honestly, we looked around at our own team and despite being really good at connectors thanks to our MCP work, we had no agents running to help us build the business.
We paused development work for a cycle (2 weeks) and everyone built agents and saw all the challenges with making them actually work. And that led su to what Tadata has become!
Congratulations on the launch @tori_seidenstein!
Tadata
@dietrichgebert Thanks so much!
CopilotKit
Hey team, congrats on an amazing launch!
Tadata
@nathan_tarbert Thank you!
Congratulations on the launch @tori_seidenstein
Tadata
@abhishek_ambad Thank you!
congrats on the launch 🙌
the "reads the room" part is what caught me. most slack bots just sit there until you @ them, so the fact that this one actually picks up context is a diffrent thing entirely.
question from someone who lives in slack all day...how does it handle channels where the convo is half decisions and half noise? does it need to be tagged or is it listening in the background?
also curious how deep the 'learns your company' part goes after a week vs after a month.
Tadata
@bencohen When you are DMing with Tadata, you don't need to tag it explictly. When you are in public channels and you want it to act, you do need to tag it. That also means channels that are a mix of noise and decisions are totally fine.
Learning happens from the very first interaction when it learns your company and your role (based on what you tell it). We see it start to provide helpful connecting the dots comments within the first week. But it grows over time of course.
Clockwork
Really like the restraint here. An AI employee that knows the company context, has access to the right tools, and knows when not to speak feels much more useful than another chatbot living in Slack. Congrats on the launch!
Dial
the delegation pitch is appealing but what's the failure mode - if Tadata researches a lead and gets something confidently wrong (stale data, wrong company, a hallucinated detail), does it flag its own uncertainty, or does that just show up as a normal Slack message and get treated as ground truth by whoever asked?
@galdayan Great question! If Tadata isn't certain it will ask the user and remember the preferences, when the answer is more complex than a simple slack message it'll create a deliverable like an attachment or a google doc/sheet where it can also detail things like process and level of confidence. like any AI agent, it can still be confidently wrong from time to time, we encourage users to read and validate results, just like you would with your human employees