I asked 10 businesses how they ACTUALLY use AI

we're probably booking somewhere around you know 100 extra 100K of AR a month when I heard the amount of money that zapier makes with AI I started to get a warm tingly feeling have you noticed that there's a ton of hype around new AI models but where's the love for businesses that are actually seeing impact from AI I wanted more stories like zapir so I sent out a tweet asking who's actually seeing real value from AI I got a bunch of responses and I interviewed 15 different people about their workflows I didn't expect to hear

so many cool applications that other businesses to take advantage of my name is Greg camad and these are my favorite ways that I've heard of AI making an actual impact hello good people you see there's a ton of new hype around new models you don't need to look very far to see more academic papers but where's the stories about AI making actual business impact I didn't want to hear from AI companies Building Products you'll just get a bunch of self-promotion I wanted to hear from the people who are actually getting value you see i' love

these stories for two reasons one I get a warm tingly feeling when someone says that a product actually gives them value and two one of the ways to spot business ideas is to notice what do people build themselves if you can make a better product and serve it at scale well you might have a business opportunity at your hands if you have an AI tool or workflow that's bringing you value I would love to hear about it links in the description so let's get into them I first chatted with Rex Harris Chief product officer at

bright they do restorative sleep powered by AI Rex and his team were swamped with survey results and chat messages from their users well enter anthropics claw 100K context window Rex and his team now use Claude as a correlation machine they have Claude connect the dots across survey results Rex is estimating that this saves his team at least 20 hours per survey in another workflow they also get a ton of messages from their customers through their inapp chat it takes a long time for human to go through them and categorize them one by one they also

use caud to look through the chat sessions and extract data like feature requests and bugs I particularly like this one because instead of begging your users for feedback with a $50 Amazon gift card you now get them in the moment passionately talking to you about your product do you have templates in your product maybe you have notion Docs Google sheet projects or code templates well creating them manually can be a big pain in the butt Dion from TASC told me that they're using AI for automated template generation before they could only do two to three

templates per week now they're doing over a 100 that's a 30 5x Improvement using AI not only is this a better experience for their customers but it also increases organic traffic through SEO does your support Channel have the same questions coming in over and over yeah well you and a thousand other businesses wouldn't it be nice if you could have ai draft a response based off of previous email answers Dion at TASC told me again that his team built a small automation that interprets new support emails does semantic search to find which emails have the

same question already answered and then drafts up a response for his team to review with this process Dion isn't letting AI run wild there's a human in Loop which is still the right way to go in my opinion Dion says that it used to take his team 3 to 30 minutes to respond to an email now it only takes him 30 seconds that's a 6 to 60x improvement with AI so awesome next is one of my favorites I got on a call with Mark Erdman of pulley who told me how his team is using slack

as a work recapper sometimes it's hard remembering what you've done at work well Mark and his team have built a tool that searches for the instances of you mentioning your own work on Slack then it Recaps it for you so not only can you evangelize yourself but you can also track your career progress and get this Mark and his team made the bot encouraging as well so not only do you get a recap of your work but you get a little pickme up at the same time we estimated this saves pulley team members at least

1 hour of work per quarter per person and it looks like pulley isn't the only one either I've been getting mentions of other groups doing this too next is the story that started it all as I was interviewing Mike Noob co-founder of zapier he mentioned that his team uses zapier with a to extract better information from sales calls then transfer over to HubSpot I caught up with Reed lead PM for AI at zapier and here's what he had to say probably one of our is probably the most impactful from a revenue standpoint um and this

is about helping our sales team just focus on like efficiency and get to conversions um and we so this one is a rather lengthy zap but it's cool because it takes the like call transcripts that happen through gong it uses open AI to extract a lot of information from those uh including like you know details from the summary recommended next steps uh adds that into back to the CRM and kind of like prods the sales rep for like how they should continue working that deal zapier estimates that this earns them an extra massive 100K of

ARR per month I caught up with Nick and Caleb co-founders of mendable their company's all about answering technical questions with AI they have a feed of all questions that their own users ask about their documentation they noticed that people are asking questions where their documentation was lacking so they built a mini bot this bot created a slack thread with the question then Nick and Caleb were asked hey is there any tribal knowledge that could help with answering this question the bot wrote draft documentation searched GitHub found any relevant files and then submitted a new documentation

PR for the team to review they say that this saves them at least 4 hours per week of writing revisiting maintaining their own documentation I love this one because it turns documentation maintenance from a reactive process to a proactive one next up is another one from zapier they noticed that 76% of users only use zapier for one use case Emily Johnson demand generation manager at zapier was looking for a way to give better onboarding recommendations to their users she built a zap that sends emails with zap suggestions for new people based off of their title

role and Company they saw their new user booking rate increase from 8% to 1.2% that's a 50% increase super cool last up I want to tell you about my favorite use case that I heard during the interviews Market pulley again told me about another slack tool that him and his team built this time it's the knowledge bot they ask this bot a question from any channel and then it answers with the tribal knowledge that was previously locked away within chat messages in slack history now I know what you're thinking but Greg slack conversations are messy

and you're right but language models in AR are great with dealing with Messy unstructured data the secret is pre-processing the raw chat messages before they go to the language model this is an important tool because it frees up people who have all the tribal knowledge AKA leadership from needing to answer questions that they've already done Mark says that his team of 30 people use this tool between 30 and 40 times a day that is an insane amount we've estimated that it saved them a whopping 40 hours of leadership time since they've built it I love

this idea so much that I wanted to find out more I sent out another tweet asking who's building this product I collected over 31 tools that I'm going to be sharing with email subscribers if you want to join the small group of people that get updates from me I can send you over the list as well links in the description I loveing doing these interviews and I want more ideas if you have an internal tool or AI product that making real impact within your company please let me know about it links in the description we'll

see you later