Men are 22% more likely to use AI daily at work and and 23% more likely to be encouraged by their managers to use it, according to Lean In research.
It is obvious by now that AI is fundamentally changing who gets leverage at work.
At Tadata we are tackling this issue head on.
Next Tuesday, Tadata's very own Tori Seidenstein will host a hands-on workshop where women will build an AI personal assistant that handles meeting prep, follow-up, monitoring, and the small tasks that quietly consume the week.
This isn’t just a conversation about what AI could do. Attendees will leave with an assistant already working for them.
Save your spot!
https://t.co/LBApuT36c5
A customer started paying us for a reason I didn’t expect: Tadata’s Google Sheets connector.
She was using AI to manage conference planning across a bunch of spreadsheets.
With Tadata, she could say “Merge these three sheets.” and then continue int the conversation to remove duplicates and reorganize the existing sheet. Then find more contacts for the conferences and add them to the sheet.
Then she drafted emails to the contacts she found asking about sponsorship opportunities, sending out via Gmail.
Importantly, Tadata kept working inside the same living Google Sheet her team already used.
That sounds so small, but it’s not possible in the mainstream Google Sheet connector.
Lots of AI products advertise their connector breadth. But saying “we have 100 connectors” doesn’t tell you how much you can accomplish with those connectors.
Can it edit the existing Google Sheet, not just create a new one?
Can it write notes back to HubSpot, not just read CRM data?
Can its web search reach the weird, hard-to-access pages on Amazon where the useful information actually lives?
All of these have been real “wow” moments for Tadata users and a reminder for our team of how much connector quality and capability really matters.
And if your team lives in Google Sheets, give Tadata one of the messy ones and tell it what you want done!
Doing a high volume of personalized outbound emails can be graceful. Yes, graceful.
Typically, outbound feels like an assembly line, where you find leads in one tool, enrich them in another, and pipe them into yet another tool for sending.
Here’s how I approach it, all in one tool.
Good marketing channels are systematic. Speaking opportunities are treated as serendipity.
So I built an event pipeline with Tadata: it finds and evaluates SF events, identifies organizers, and drafts personalized outreach. I decide which ones to pursue.
Here’s how it works.
We spent a month preparing for Product Hunt, earned 343 upvotes, doubled our website traffic, and still came in #2.
Here’s what worked and what we’d do differently:
1. Start ~1 month before launch day.
Product Hunt seems like a one-day competition, but it’s really an operations project. Good products rise to the top, but a good product with poor launch execution will flop.
One month in advance, we held a team meeting to create our game plan. That turned into 49 Linear issues, which ranged from creating our launch video to load testing the product.
2. Build your product story across your assets.
We made a video (linked in a comment) with a funny 20 second hook followed by a 1 minute, high-level product demo. Plus static images, a tagline, and maker comment.
3. Get reviews for customer proof.
We asked top users to review Tadata.
Our biggest mistake was this: we thought Reviews should be published on launch day. We learned that Product Hunt only publishes a review if the reviewer’s profile is complete. Several of our customers had new or incomplete profiles, so their reviews never appeared.
4. Map our the supporters in your orbit.
Each person on our team created a list of customers, friends, former colleagues, investors, and relevant Slack and WhatsApp communities, with different ways for each to engage.
We tailored the outreach based on the relationship. Customers could talk about the product. Friends could extend our reach. Communities could introduce us to the right audience.
5. Build Product Hunt relationships.
Every morning, Tadata briefed us on a few relevant Product Hunt launches. Someone on our team would try the product, leave thoughtful feedback, and connect with the maker. When the time came, we asked them to check out our launch.
6. On launch day, pay attention to what’s working.
We posted across our social channels, contacted our networks, answered every Product Hunt question, and helped new users through any issues.
Tadata also flagged to us that one co-founder’s Tweet was outperforming the rest. We directed more people toward it, which helped extend its reach.
7. Explain Product Hunt basics to people.
Another oversight. Some people thought they had supported us while logged out because the visible counter increased. Their votes didn’t count. Others didn’t realize that Product of the Day is decided within that specific 24-hour window. Don’t assume your audience understands how Product Hunt works; overexplain.
8. Ignore the paid-vote offers.
We received a shocking number of them. We ignored them.
Our final results:
#2 Product of the Day
343 upvotes
40 substantive comments
2× website traffic and a meaningful increase in daily signups
And we continue to see elevated traffic and sign-ups after launch day.
🏆 #2 Product of the Day on Product Hunt
380+ upvotes, 40+ comments, and the Product Hunt daily newsletter. On a Sunday. Thank you to everyone who showed up.
Until two days ago, the way to hear about Tadata was talking to one of us in a GTM Slack community or at a meetup in downtown SF.
Now it's out in the world: an AI employee that lives in your Slack. It arrives with context, notices the work worth doing, and offers to take it. You don't build it or configure it. You just work.
Get in on the hype: https://t.co/08pehW3zXR
When I tell candidates we’re building an AI employee in Slack, they hear the form factor and say there are already AI employees in Slack. fair.
My answer is that there are already many people out there too, and I still care a lot which ones I work with. Often, the heart of an idea isn’t in its title.
Btw, we’re live on Product Hunt now with AI employee in Slack that is fun to talk with! and would really appreciate your support:
https://t.co/mPyO9LYXtF
Tadata is live on Product Hunt today. Come support us with an upvote/comment ❤️
https://t.co/UrlrM58hKs
Using AI at work should feel like delegating to a great teammate, not learning how to build and manage agents.
Tadata reads the room.
My co-founders and I met back in 2017, but starting a company together only happened recently.
Miki and I met in a startup accelerator in Tel Aviv. The program brought 10 Americans and 10 Israelis from the intelligence unit, and threw us in a room for four months basically saying, “go start a company!”
Itay was a childhood friend of someone else in the program, and he and Miki later became university classmates.
We didn’t end up pursuing the ideas we explored that summer. We went off and did other things.
Miki ended up at Cyera, focused, as he says, on preventing people who shouldn’t have access to data from having access. Now, he’s flipping that around to ensure that people who should have access to data to do their work do.
Itay started working on these “generative model” things, years before anyone knew that was cool.
And I built my first startup, FairStreet, getting about as much firsthand education in building a company as I could have asked for.
In retrospect, Tadata makes a lot of sense given what each of us spent those seven years doing.
We didn’t start from a blank slate. We kept in touch, and when we finally decided to build together, each of us brought seven years of life experience with us.
Our team had 3x more chats with our AI agent this month.
What changed? We consolidated our Tadata agent into a single agent that spans everything from sales to product use cases. Second, we embedded that unified agent into our team Slack.
Then we watched usage skyrocket for ourselves and our users, for several reasons:
1. 𝗗𝗲𝗹𝗲𝗴𝗮𝘁𝗶𝗻𝗴 𝘄𝗼𝗿𝗸 𝘁𝗼 𝘁𝗵𝗲 𝗮𝗴𝗲𝗻𝘁 𝗯𝗲𝗰𝗮𝗺𝗲 𝗹𝗼𝘄𝗲𝗿-𝗳𝗿𝗶𝗰𝘁𝗶𝗼𝗻. This one's obvious. Now we can fire off tasks in the tool we already live in for communication. And no more digging through lists of agents or projects; we say what we want done, and Tadata routes it correctly.
2. 𝗨𝘀𝗲 𝗰𝗮𝘀𝗲𝘀 𝗯𝗲𝗰𝗼𝗺𝗲 𝗰𝗼𝗻𝘁𝗮𝗴𝗶𝗼𝘂𝘀. Teammates see each other use Tadata in Slack, discover new workflows, and start using them too.
3. 𝗠𝘂𝗹𝘁𝗶𝗽𝗹𝗲 𝘁𝗲𝗮𝗺𝗺𝗮𝘁𝗲𝘀 𝗰𝗮𝗻 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗲 𝗱𝗶𝗿𝗲𝗰𝘁𝗹𝘆 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗮𝗴𝗲𝗻𝘁 𝗼𝗻 𝗮 𝘁𝗮𝘀𝗸. One teammate can start a thread with the agent, then someone else can jump in with follow-up questions.
4. 𝗧𝗵𝗲 𝘀𝗵𝗮𝗿𝗲𝗱 𝗮𝗰𝗰𝘂𝗺𝘂𝗹𝗮𝘁𝗶𝗼𝗻 𝗼𝗳 𝗰𝗼𝗻𝘁𝗲𝘅𝘁 𝗸𝗲𝗲𝗽𝘀 𝘂𝘀 𝗰𝗼𝗺𝗶𝗻𝗴 𝗯𝗮𝗰𝗸. Someone on the business side taught the agent which accounts are test users while they were analyzing new user sign-ups. Later, someone on the tech side was making a product usage chart. They didn’t have to remember to say, “Oh, and make sure you filter out all our test accounts” because Tadata already knew.
The last one is a huge deal.
Every company accumulates tribal knowledge like:
“When we say activated, we mean X.”
“This field stopped being reliable after March.”
“Exclude these accounts from analysis.”
Getting that context out of individual people’s heads and into something the whole team can reliably use has always been hard (and it was a big topic during my time at Facebook!)
Our bet with one shared Tadata agent in Slack is that busy people won’t build and organize separate agents or curate their knowledge. They’ll just do their work, and Tadata can organize the context for the team as they go.
So the chart might look like a story about less friction increasing usage after moving Tadata into Slack.
It’s really about how much more work an agent can handle once it becomes shared infrastructure for an entire team.
Unfortunately the most interesting part of speaking at AI Tinkerers GTM last night was not my talk.
I got two questions afterwards that were better.
The talk was about something that looks simple and isn’t: call prep. I used it to talk about how you actually build a no-slop agent.
Just telling the model “be brief” fails miserably. You walk into the call missing the one detail that would have made you sound like you’d been following the account for months.
Call prep isn’t one task. You’re stitching:
- how you got connected (Gmail, LinkedIn)
- what was said last time (CRM, Granola)
- context on the company and person (web)
What worked for us: keep the same harness, swap in the right model at the communication step.
Then the Q&A. Someone asked how far Tadata can actually reach into HubSpot today.
Answer: further than most tools, including HubSpot’s official MCP, because we grew up building MCP servers in the open.
Someone else asked the ultimate one: how do you measure the ROI of extraordinarily good call prep?
Answer: Tadata is system-led, so the brief is not a one-off artifact. You use it to walk into the meeting. You also use it at the end of the week — pull the transcripts, look at which meetings converted, and ask what actually correlated. Segment research. Hiring signals. Last-meeting objections. Something else entirely.
Both questions were getting at the same thing: how deeply is the AI embedded in the system? Deep enough to get the right context, take useful action, and learn whether that action worked.
Great call prep is useful in the moment. Tadata gets better at it because it can see which prep leads to the outcomes.
That’s the part I’m most excited about: the self-improving flywheel!
Last night @ToriSeidenstein was on stage at AI Tinkerers' GTM Engineer Night in San Francisco.
Her point: answering a question in two lines is a harder call than answering in twenty. Restraint is the actual skill in a GTM agent, not range.
The room took it somewhere better. Less "can it answer" and more "how far into the real system can it sit before it has to."
That's the version of this conversation we show up for.
📦 Just dropped 📦
You can sign into a new connector straight from Slack now, no separate tab, no app switch [:chefs-kiss:]
One more thing! recurring job tells you why it's messaging you, not just what it found, so a scheduled update doesn't read like it came out of nowhere 🎉
We know.. none of these are the exciting kind of ship note. BUT, they're the kind that decides whether you trust the next nine things it tells you [:we-smart:] 🤓
Free to start, $50 in credits:
https://t.co/PxpRmKCnOw
@ToriSeidenstein with yet another angle that makes Tadata the perfect tool for forward looking sales people.
The insurance verification example is the whole thing: a clinic complaining about coverage confusion on Google Maps is a warmer lead than a Series A announcement. @tadata_team is built to sit on signals like that and hand you the account before anyone else notices it.
I think Google Maps is one of the most underrated sources of sales signals.
Most companies are watching the same things: fundraises, hiring, job changes.
They're useful, but crowded. Every seller reaches out, the prospect is flooded with inbound messages, and nobody stands out.
The alpha is in signals specific to what you sell.
One Tadata customer sells insurance verification software to clinics. Their Tadata monitors Google Maps reviews for complaints about insurance eligibility.
A patient writes: “Front desk couldn’t confirm my coverage, had to reschedule twice.”
That’s a unique buying signal for the pain point this company solves.
And the interesting signals can live anywhere:
Yelp reviews.
G2 reviews.
LinkedIn posts.
Press releases.
Product release notes.
Tadata can help you figure out which weird signals matter for your business, then monitor them and bring the relevant leads into Slack.
Tell Tadata what you sell and ask:
“What signals should I be watching that my competitors probably aren’t?”
Our CEO, @ToriSeidenstein at Inkle's GTM Engineering meetup with Clay, Rho and Emergent, sharing some thoughts on the relationship between sales and GTM. We at Tadata aim to empower every Seller with GTM capabilities.
Was at the GTM Engineering meetup last night hosted by Clay (which helped coin the term), and kept thinking about this thread.
The shift isn’t “sales” → “GTM.” It’s that a lot of sales work is becoming programmable.
The winning model IMO: centralized data for things like lead lists + memory, and every rep is able craft customized tools on top of that which let them go out and just sell
Was at the GTM Engineering meetup last night hosted by Clay (which helped coin the term), and kept thinking about this thread.
The shift isn’t “sales” → “GTM.” It’s that a lot of sales work is becoming programmable.
The winning model IMO: centralized data for things like lead lists + memory, and every rep is able craft customized tools on top of that which let them go out and just sell