In this episode of AI First Principles, Brian speaks with Sid Bharath about why mid-market companies are falling behind larger enterprises and AI-native startups—and what leaders can do about it. Sid explains why adopting AI is both a strategy and execution problem, especially for companies still relying on disconnected software, repetitive administrative work, and employees manually moving information between systems.

The conversation explores agentic AI, AI employees, executive AI chiefs of staff, workflow automation, and the measurable ROI that comes from returning time to high-value work. Sid also explains what it means to “re-found” a business with AI at the core rather than simply layering new technology onto old processes.

The takeaway is clear: becoming AI-first does not require starting from scratch. It requires leaders to rethink how work gets done, experiment personally with AI agents, and identify the friction that technology can remove.

Chapters:
(0:39) Meet Sid and the mid-market AI opportunity
(0:53) Why mid-market companies are falling behind
(2:44) AI strategy versus execution problems
(3:16) Moving beyond ChatGPT into agentic workflows
(4:23) What it means to re-found a company with AI
(6:43) Finding friction inside disconnected business tools
(8:28) Measuring the ROI of AI automation
(9:53) Finding fast AI wins inside a company
(10:22) Building an AI chief of staff
(11:47) Replacing app switching with one AI interface
(12:26) Connecting personal and professional workflows
(14:07) Lessons from building multiple AI companies
(16:20) Why some AI business ideas work
(17:36) What AI-first mid-market companies will look like
(18:26) The AI shift leaders must make now
(19:33) The AI habit every executive should build
(20:21) What companies overcomplicate about AI
(21:19) Sid’s advice for building an AI-first company
(22:09) How to connect with Sid

Sid

(0:00) I think that the core shift right now is that they themselves, as leaders of a company, should be the ones championing the use of AI and understanding how to change the way they work to an agentic way of working. (0:17) So they need to be doing things like having their own AI chief of staff, like we talked about earlier, or changing the way they work and really playing around with AI tools and agentic AI. (0:29) And building your own AI agents are getting something like Cloud Co-Work is a good example that you don’t have to build it yourself as much as you shape it or mold it.

Bryan

(0:39) Welcome to the AI First Principles podcast. (0:42) We have Mr. Sid Bharath. (0:45) Did I get that right?

(0:46) Here with us today.

Sid

(0:47) Yes, right. (0:48) Thank you so much for having me.

Bryan

(0:50) Yeah, you’re welcome.

Sid

(0:51) Thank you. (0:51) Thank you for having me.

Bryan

(0:53) Hey, so when people talk about AI, it’s usually big tech or early stage startups. (1:00) Sounds like you’re mainly focused on mid-market. (1:03) So where do mid-market companies get stuck and how are they different than the others?

Sid

(1:07) Yeah, I mean, it’s interesting. (1:09) You know, I think in this wave of technology, the enterprise companies have been very fast adopters, primarily because a lot of these enterprise companies have already had internal teams working on data and internal IT teams. (1:26) So to be able to adopt AI was a lot easier for them because they already had some capacity and people in there who could just use these AI tools on their data.

(1:37) And then with tech startups, of course, they’re always at the cutting edge of tech. (1:41) A lot of the new age companies that are coming out of these accelerators like Y Combinator are AI-first companies. (1:49) But the mid-market, like the small to mid-sized businesses, they’re like your traditional, you know, you go to take either, say, you know, your law firms or your mom and pop stores to your, even your like big brand name TTC companies and manufacturing companies, companies like that.

(2:05) They tend to not have that kind of like internal talent and they tend to not move as fast and like new tech. (2:15) And now, of course, they want to catch up and they want to like really implement AI and start to get the benefits that every other company is seeing. (2:22) And so that’s kind of where we are focusing on because there’s like a huge demand there and not a lot, they don’t have the talent already to do it themselves.

Bryan

(2:30) Ready to lead smarter and invest wiser? (2:33) On the Executive Connect podcast, we unpack executive strategies for wealth and influence. (2:40) Hit the subscribe button now.

(2:42) Don’t just watch, act. (2:44) Okay. (2:45) So do you see that as more of a strategy problem or is the strategy there and it’s like an execution problem once they decide that they’re going to move?

Sid

(2:53) I think it’s a bit of both. (2:54) And I think a lot of times they don’t have either the strategy or the capacity to execute. (2:59) Right.

(3:00) And so, you know, when you ask a company, like, what are you doing with AI? (3:06) A lot of them just, they don’t really know what I can do as beyond using ChatGPT, you know, to ask questions and get answers. (3:15) Right.

(3:16) It’s funny that like a lot of people are still stuck on just ChatGP as a chatbot that you could use to give you like a Q&A bot. (3:24) But beyond that, getting AI to do work for you, like agentic AI and building agents that can actually do tasks. (3:32) So you think about, you know, let’s take an example of a sales team and a lot of work in sales is it’s not just like getting on a call.

(3:40) It’s like you have to update your CRM and you have to create proposals and presentations and, you know, feedback and so on. (3:48) A lot of that is kind of like work that the salesperson is doing that they could hand off to an AI agent to do and focus on doing the sales calls. (3:54) You take any team in a company, there’s a lot of admin related work that people are doing where, you know, even if they’re using AI right now, they’re using ChatGPT to answer them, answer questions for them, but they’re not using it in the way that they should be using like an agentic flow.

(4:11) And so I think there’s like, A, the strategy of figuring out how your team should be using AI and B, building out those agentic workflows for them to use it the right way.

Bryan

(4:23) So you talk a lot about helping companies re-found themselves with AI at the core. (4:30) What does that look like in practice?

Sid

(4:32) Yeah, you know, I’ll tell you about our company as well. (4:37) Like, you know, we’re running a services business, right, essentially, and we’re working with dozens of brands and big name companies and we have a small team. (4:44) And how are we able to like service these big companies with the small team is because we have started off with AI at the core.

(4:51) And so re-founding with AI at the core means moving from the way you used to work of everyone is doing these manual repetitive workflows or admin tasks to handing that over to an agent and changing the way that you are working essentially. (5:08) So instead of your people doing all of these admin tasks, you now have AI agents as employees that your people work with and manage who do the tasks for you. (5:21) And that is a completely different way of working than you have before.

(5:26) So I think like if you think about, has, if anyone listening to this is thinking, well, the way I worked is pretty much the same as it was six months ago or a year ago or two years ago, then something’s wrong. (5:37) Because for me, the way I work is completely different than I worked three months or six months ago. (5:41) I don’t log into different dashboards and tools.

(5:45) I have my personal AI agent that I talk to every day and that AI agent does work for me, but I’m able to think about more strategic and creative stuff and have the agent execute work on my behalf. (5:58) And I think that’s what re-founding means is to like start from like build the business again, but with AI at the core.

Bryan

(6:07) Money Ripples is on a mission to help professionals like you get their money working harder. (6:14) Their clients free up an average of $35,000 their first year without having to work extra hours. (6:22) To show you how they’ve put together a powerful training called cashflow secrets.

(6:27) And as a listener of the executive connect podcast, you can get it completely free. (6:33) Just visit moneyripples.com forward slash secrets and enter the promo code E-X-E-C. (6:43) So when you walk into a company that you’re proposing to do this work for, what’s the first thing you look at to determine whether you’re ready for the shift or not?

(6:52) Are there some common patterns there?

Sid

(6:54) Yeah, there are. (6:55) I think like a lot of times you’ll have companies that have all these disconnected tools. (7:02) You’ll take a sales team example, right?

(7:05) Like you have your sales force and then you have your outreach tool and then you have your transcript tool and then you have your documents and there’s five or six or 10 different tools for each team. (7:19) Marketing team has like dozens more platforms like your ad platforms and your spreadsheets and so on. (7:25) And nothing is talking to each other.

(7:27) And so that’s kind of like a very common pattern to see when you come in and like nothing is talking to each other. (7:33) And it’s a human who has to take data from one tool to another. (7:36) And it’s a waste of your time and your talent to be doing that.

(7:40) If you’re a marketer or a salesperson, you shouldn’t be copying and pasting stuff and downloading stuff and uploading things over and over again. (7:47) Or you should not be opening multiple dashboards. (7:50) And so that’s a very common pattern.

(7:53) But it’s not a problem. (7:54) It’s fixable with AI. (7:56) The first thing we do is we understand okay, what tools are you using right now?

(8:01) What is your workflow? (8:02) How does your day-to-day look like? (8:05) And try to figure out where are you spending a lot of time doing admin work, copying, pasting data, that kind of stuff.

(8:10) That’s how we know okay, those are the places of friction. (8:14) That’s where we can build AI agents to do that work for you. (8:18) And you can then get back time to focus on what you’re meant to be doing, getting those sales calls, creating creative marketing campaigns, that kind of thing.

Bryan

(8:28) So for a lot of companies at a high level, the big question is simple. (8:32) What’s my return on the investment that we’re going to be making here? (8:35) So how do you approach that?

Sid

(8:36) There’s multiple ways. (8:38) I think the number one, the easiest way to calculate that is just to look at how much time is being spent by people doing admin work or manual work, copying, pasting data. (8:50) And that’s easily calculatable, right?

(8:51) We spoke to a sales team the other day where they estimated that each sales rep was spending two hours of their day. (8:58) So two hours out of their eight-hour workday, it’s 25% of the day in admin-related tasks, right? (9:05) So if your salesperson is able to only spend six hours out of eight on sales calls, there’s a 25% drop in the number of sales calls you could be doing.

(9:15) That’s a 25% in potential revenue drop, right? (9:20) So it’s very easy to calculate that kind of stuff where you can say, okay, my people are only effective 75% of the time or 60% of the time because the rest of the time they’re doing admin-related work. (9:34) And every company has a metric for output per person or output per employee or revenue per employee.

(9:40) And now the employee could then become 25% or 30% more productive. (9:45) Then that revenue per employee goes up, right, by 25 or 30%. (9:49) So there’s a very clear line to ROI from that.

Bryan

(9:53) So do you usually find quick wins right away whenever you’re dealing with companies or does it usually take time before the impact becomes clear?

Sid

(10:01) Oh, it’s very fast. (10:02) Like we can, sometimes some companies already understand like where the big roadblocks are and the bottlenecks, but otherwise it takes us a few conversations with people. (10:11) Like we’d say, we’d start with a certain department in a company, talk to a few people there.

(10:16) And then within a couple of days, we already have like dozens of ideas of where we can create quick wins.

Bryan

(10:22) So you’ve been talking about this idea of an AI chief of staff. (10:26) What problem is that really solving for executives?

Sid

(10:29) So I have this kind of, I have my own chief of staff, right? (10:32) They call it Jarvis, not a very creative name. (10:34) But my chief of staff here is essentially, I wake up in the morning and it has my morning brief ready.

(10:41) So it’s looked through my emails, it’s looked through my calendars. (10:43) Like I said, you have a podcast coming up with Brian, your prep notes. (10:47) It’s already looked at the Google Doc you shared from your email and it said, here are some prep notes.

(10:52) You have a conference coming up and I’ve looked to the attendees of the conference and you should meet X, Y, and Z person, right? (10:59) Oh, by the way, you also got an email from John and they want you to send a proposal for phase two of the project. (11:06) I’ve looked through the project details, I’ve taken the proposal template, drafted up a new proposal.

(11:12) Here, do you want to send this to John, right? (11:14) And so it’s gone ahead and it’s managed my day for me and all I’m doing is like, yeah, okay, the proposal looks good or change it, change to that. (11:23) Okay, yeah, let me read who I should meet at this event and go meet those people.

(11:28) And that’s the value there of me never having to open up my email and calendar and my company dashboards and CRM and so on, where everything is just done for me in one go. (11:42) And maybe I’ll be able to get work done without me having to do all of that stuff.

Bryan

(11:47) Yeah, so instead of jumping between apps, you’re basically working through one interface to do all this work.

Sid

(11:54) Exactly. (11:55) And I’m just chatting to my agent, right? (11:57) And I could chat with it on my computer or my mobile phone.

(12:00) If I’m on the go at a conference, I can open up the thing and text it and be like, hey, find me someone else to meet. (12:07) I’ve met five people you suggested, right? (12:10) It’s that kind of thing, where it’s like having a real chief of staff with you, but it’s all AI.

(12:18) It has access to your data, your tools, your business context, and it’s able to make the right calls because of that.

Bryan

(12:26) Yeah, I know a lot of folks have side hustles they’re doing and they’ve got their personal life and all of that. (12:33) How do you handle the split between work and personal life? (12:38) Does that sometimes get messy or does the agent know where the lines are?

Sid

(12:43) I’ve made mine, I’ve built my personal life into my agent. (12:46) So my agent does know how the lines work and have created like sort of, okay, here’s the business, here’s the tools from the business, and here’s my personal life, and here’s what my goals are outside of my business, right? (12:59) I mean, I think in my case, because I’m a business owner, they’re obviously intricately connected, but the agent does know, okay, like, hey, you have your personal training session at 11 o’clock, and by the way, you spent like 10 hours a day working today.

(13:17) Have you taken a break or have you gone and spent some time with some friends or family or have you called up a friend? (13:23) So it is able to manage that because it’s able to keep track of my day by looking at my email, my personal email, and the calendars.

Bryan

(13:32) Today’s episode is brought to you by Renaissance, redefining success for modern Mavericks, a book I co-authored with Michelle Kine. (13:42) It’s a straightforward guide through reinvention, leadership, and building real freedom. (13:48) Inside, you’ll find practical frameworks, the Maverick test, and stories that help you take action today.

(13:55) Pick it up at executiveconnectexperience.com forward slash Maverick. (14:01) Use the code EXEC for 10% off. (14:07) So you were previously at Forum Ventures, and you were building multiple AI companies every year.

(14:12) What patterns stood out as part of that experience?

Sid

(14:16) Yeah, I think like, you know, the crazy thing is like when I was at Forum Ventures, maybe it was like I left last year or so, you know, the year before that, and the year prior to that. (14:26) So like about two and a half years, and we were building these AI companies. (14:29) And I think like the way that companies are being built have already changed.

(14:34) That’s how fast the AI space is moving, right? (14:36) Like, I think if I was back there, I would be working in a completely different way and doing completely different things. (14:43) But, you know, to your question back then, I think the core patterns were we would come up with, say, an idea or something, and we would talk to companies, and we would kind of understand, it’s like, is this a problem that you have?

(14:58) Do you, you know, have this kind of pain point? (15:02) Like, can we dig into that a bit more? (15:04) And it was really just like being able to build custom AI solutions or AI solutions for a company and seeing if multiple companies had the same pain point, and we could build that one AI solution for them, right?

(15:17) And I mean, I think that’s kind of like, in general, good startup advice or business building advice, right? (15:23) Can you find a pain point that businesses have and there are enough businesses that have that pain point where you can build a product? (15:31) Now, of course, like I said, AI has changed so fast and things have changed so fast, the way you build a business has changed so fast that now a lot of VCs are also saying that, hey, a services company can also be as scalable as a product company, as a software company, because AI allows you to build so fast, right?

(15:52) So, I can use my agent to go and build things, a custom version for a client as fast as, say, writing an email, right? (16:01) It’s just like, it’s so fast, it can build that for a client, and then I can just duplicate that for another client and customize that, and it’s done quickly, right? (16:08) And so, you know, there’s this, we’re moving to a world where a service, customized service, can be done at the scale and speed of a pre-built product, and I think that’s exciting.

Bryan

(16:20) What separates the ideas that actually work from the ones that don’t?

Sid

(16:25) That’s a good question, you know, and I think one of the things is, like, it goes back to what I said previously about really trying to understand what is that pain point that the customer has and whether enough customers have that pain point, right? (16:39) And so, you know, if you’re trying to solve a problem that doesn’t really exist or that not too many people have, that’s where you’re gonna, you’re gonna fail, right? (16:47) So, you really need to have some sort of deep (16:49) understanding in a market or an industry or a certain space where you, like, I know this is a (16:56) problem, I felt the pain, you know, I’ve talked to, like, 15, 20 people, they’ve all felt the pain, (17:03) and, you know, just by doing research online, I can see that many other people are feeling this (17:07) pain too, and that’s where you can, you know, you can build something successful. (17:12) But if you don’t do that, if you skip through all that, and I myself have done in the past, I’ve been guilty of this, of skipping through those steps because it feels like it’s not work or it feels like it’s, you know, like, if I felt the pain that I’m sure there must be a thousand other people feeling this, and I’ll just skip through it and go, but that’s not the thing, right? (17:30) Like, you could easily mistake it for something that’s just maybe unique to you and not to anyone else.

Bryan

(17:36) If we look years ahead, you know, it’s changing quite rapidly, right? (17:40) What do mid-market companies that get AI right look like compared to ones that don’t?

Sid

(17:45) I think the ones that are getting it right are the companies that are really understanding that this requires a shift in the way that you work, and that AI is not just another tool that you plug in, right, or just like a chatbot, and that there’s more to it. (18:07) It’s like, it’s turning into an agent. (18:09) It’s capable of doing tasks and taking instructions and thinking and behaving like an employee.

(18:17) And so, unless you start to utilize AI like it is a sort of employee, you know, you’re not going to be getting the full potential of it.

Bryan

(18:26) What’s one shift leaders should be making now that most are still overlooking?

Sid

(18:30) I think that the core shift right now is that they themselves, as leaders of a company, should be the ones championing the use of AI and understanding how to change the way they work to an agentic way of working. (18:47) So, they need to be doing things like having their own AI chief of staff, like we talked about earlier, or changing the way they work and really playing with AI tools and agentic AI and building their own AI agents are getting something like, you know, cloud co-work is a good example that you don’t have to build it yourself as much as you shape it or mold it. (19:10) So, you don’t need to be technical even.

(19:12) You can create this AI chief of staff using cloud co-work to do things for you, but you have to explore. (19:17) You have to really go and try it out yourself as a leader of a company before you can expect the rest of your company to follow and to work the way that you’re working, right? (19:25) And you just share that stuff and you have to be like the leader who is doing things and people can follow the way you’re doing it.

Bryan

(19:33) Let’s close with some rapid fire. (19:35) What’s one AI habit every executive should be building today?

Sid

(19:39) I think the number one thing is like you look at, you think carefully about the work that you do and the stuff that you spend so much time on. (19:45) I know every executive is drowning in emails and calendars and meetings and so on, right? (19:50) I myself was doing that, running a small business myself.

(19:53) I can’t imagine what it must be for an executive at a bigger company. (19:58) And so, try to figure out where can you hand off work to an AI agent, right? (20:03) Or maybe if you’ve, a good way to frame it is if you had an AI assistant or an assistant, a human assistant today, what would you hand off to them?

(20:11) And instead, try to use AI to do that. (20:14) And I think that you just start there, start with something small, experiment and then see where it goes from there.

Bryan

(20:21) So what’s something about AI most people overcomplicate?

Sid

(20:24) I think one thing is like everyone thinks that you have to be prepared or you have to get all the ducks in a row and maybe you have to have the right data and all of that. (20:33) But honestly, you don’t. (20:38) It’s not like you’re training a new model from scratch or anything.

(20:41) You’re using existing models to connect to your source of data. (20:45) And if you don’t have a centralized database, that’s fine too. (20:47) You could just connect it directly to your various tools.

(20:51) And AI agents are smart enough to figure out which tool to use when, if you can build it the right way or prompt it the right way. (20:59) And so, I think you don’t have to say, okay, we’re not ready for it yet. (21:08) Or it’s not like we don’t have the right data or whatever.

(21:10) No, it’s just start using it, connecting it to your tools, seeing what happens, seeing where it breaks and then fixing it. (21:17) It’s not that difficult.

Bryan

(21:19) What’s one piece of advice for someone trying to build an AI-first company?

Sid

(21:23) I think the number one thing is, you know, understanding that the way you do work has changed. (21:32) And if you’re building an AI-first company or any company in any industry and you want to be AI-first or AI-native, then you have to change the way you work. (21:48) And you have to think of AI as your central intelligence that sits between you and your business tools and data and does work on your behalf or helps you connect tools and data as well.

(22:03) And so, if you can think of it that way, if you change the way that you’re thinking about work, then you’ll be able to use AI the right way.

Bryan

(22:09) Well, Sid, I really appreciate your time today. (22:12) Did you want to give our listeners an idea about how they can get in touch with you?

Sid

(22:15) Sure. (22:15) Yeah. (22:16) Thanks for having me on, Brian.

(22:17) This was great. (22:18) And yes, if anyone wants to get in touch with me, they can go to refoundai.com, R-E-F-O-U-N-D-A-I.com and you’ll have more information about our business and how we can help you. (22:30) And there’s some contact details there.

(22:33) And my email is sid at refoundai.com. (22:35) So you can just email me directly too.

Bryan

(22:37) All right. (22:38) Well, thanks for the time and thanks for being a part of the podcast. (22:41) And that’s the AI First Principles podcast.