Praveen Ghanta is the Co-founder and CEO of Fraction, which specializes in pairing businesses with top-notch fractional technical talent in the United States. He is a seasoned entrepreneur who founded four companies, with two successful exits, including the sale of HiddenLevers to Orion in 2021. Praveen is adept at leveraging senior expertise to enhance software development productivity and firmly believes in the potential of AI to transform the field. As an experienced bootstrap entrepreneur, he understands leveraging fractional workforces to accelerate business success.
Here’s a glimpse of what you’ll learn:
[2:34] Praveen Ghanta discusses how Fraction is revolutionizing the hiring of fractional talentÂ
[6:32] The types of professionals who join Fraction and their motivationsÂ
[11:34] Myths and misconceived notions about fractional hiring
[14:20] Insights on ensuring knowledge and talent retention in fractional hiringÂ
[16:21] Praveen shares the importance of integrating fractional talent into your team's cultureÂ
[18:14] The ideal client profile for Fraction's services
[22:58] The transformative impact of AI on software development productivity
[34:09] Lessons in leadership and the impact of positive friction on decision-making
In this episode…
When navigating the choppy waters of startup growth, many founders find themselves grappling with talent acquisition. Skilled individuals often command high salaries or are already engaged elsewhere, posing a significant hurdle. But what if there was a way to tap into this talent pool without traditional constraints?
Serial entrepreneur Praveen Ghanta reveals how fractional hiring has become the lighthouse guiding businesses to safe harbor. He draws on his own experiences with his company, HiddenLevers, to illustrate the power and flexibility of working with fractional talent. Additionally, he discusses how Fraction connects businesses with senior-level experts for part-time work, enhancing productivity and saving costs. Praveen also dives into AI's exciting role in software development, sharing insights on its ability to accelerate the engineering process when paired with experienced minds.
In this episode of The Customer Wins, Richard Walker interviews Praveen Ghanta, Co-founder and CEO of Fraction, about revolutionizing team dynamics and productivity. Praveen discusses how Fraction is revolutionizing the hiring of fractional talent, myths and misconceived notions about fractional hiring, and the importance of integrating fractional talent into your team's culture.
Resources Mentioned in this episode
Quotable Moments:
"The best talent already has a job, so why not hire them fractionally?"
"Fractional work is where the growth is, giving access to talent you couldn't otherwise afford."
"Pairing generative AI with really senior talent yields about a 4x productivity gain."Â
"When launching your MVP, that's the starting gun, not the finish line."Â
"Having the right fractional person's time can significantly impact your business success."Â
Action Steps:
Embrace fractional hiring for specialized roles: Incorporating fractional professionals enables access to high-level expertise without the full-time cost, addressing the talent acquisition challenge.
Focus on effective communication tools: Ensuring seamless remote collaboration through platforms like Slack or Teams is crucial for integrating fractional talent into existing teams.
Prioritize cultural fit in fractional partnerships: Aligning with professionals who share your company's values enhances team cohesion and effectiveness.
Leverage AI tools to augment development: Utilize AI to expedite the research and coding process to stay ahead of the tech curve.
Cultivate constructive co-founder relationships: Engage in open, constructive dialogues with co-founders to reach refined, synergistic business decisions.
Sponsor for this episode...
This is brought to you by Quik!
At Quik!, we provide forms automation and management solutions for companies seeking to maximize their potential productivity.
Using our FormXtract API, you can submit your completed forms and get clean, context-rich data that is 99.9% accurate.
Our vision is to become the leading forms automation company by making paperwork the easiest part of every transaction.
Meanwhile, our mission is to help the top firms in the financial industry raise their bottom line by streamlining the customer experience with automated, convenient solutions.
Go to www.quickforms.com to learn more, or contact us with questions at support@quikforms.com.
Episode Transcript:
Intro 0:02Â
Welcome to The Customer Wins podcast where business leaders discuss their secrets and techniques for helping their customers succeed and in turn grow their business.
Â
Richard Walker 0:16Â
Hi, I'm Rich Walker, the host of The Customer Wins, where I talk to business leaders about how they help their customers win, and how their focus on customer experience leads to growth. Some of our past guests have included Cormac Murphy of CashTech advisor solutions, Mark Friedenthal of Tolerisk, and Corey Westphal of Mobile Assistant. Today, I'm speaking with Praveen Ghanta, CEO and founder of Fraction. And today's episode is brought to you by Quik!, the leader in enterprise forms processing. When your business relies upon processing forms, don't waste your team's valuable time manually reviewing the forms.
Â
Instead, get Quik!, using our Form Xtract API, simply submit your completed forms and get back clean context, rich data that reduces manual reviews to one out of 1000 submissions. Visit quikforms.com to get started. Before I introduce today's guest, I want to give a big thank you to Michael Lieberman of Advisors Capital Management for introducing me to Praveen. Go check out their website at advisorscapital.com and find out how they tailor portfolio management solutions to meet the unique financial objectives of advisors clients.
Â
Okay, I've been looking forward to this episode for quite some time. Praveen Ghanta is the CEO and co-founder of Fraction. He's the founder of four companies. Has had successful exits on two that were bootstrapped. That's familiar to me, I'm bootstrapped. You may know him from having built HiddenLevers over a decade serving financial advisors with portfolio stress testing and proposal generation functionality. He sold HiddenLevers to Orion in 2021 and now he's building the future of work at Fraction. Praveen, welcome to The Customer Wins.
Â
Praveen Ghanta 1:51Â
Yeah, thanks for having me Rich. And I think I was mentioning this to you before, but I wish that we got to know each other better while I was at HiddenLevers, because we could have used Quik! with so many of the advisor Honestly, all the DocuSign madness, if I never had to deal with that again, and we'd known about Quik, I think that would have been very helpful.
Â
Richard Walker 2:10Â
So, man, there's so many opportunities like that. I wish, I wish, but we're here now, and hopefully we do get to work together. So for those who haven't heard this podcast before, I talk with business leaders about what they're doing to help their customers win, how they built and deliver a great customer experience, and the challenge is to growing their own company. Now, what Praveen is doing, I think, is awesome. So let's understand your business a little better. Praveen, how does your company help people?
Â
Praveen Ghanta 2:34Â
Yeah, so how do we help our customers, there's a bit of a backstory. So Fraction firstly, what do we do? We help companies to work with fractional talent. Sort of we have our tagline, if you will, is the best talent already has a job, so why not hire them fractionally? Which, I think everybody agrees that, yeah, the best folks do, typically already have a job. That part is true, what we sort of perfected over a decade at my last company, HiddenLevers, was working with fractional talent, and we were able to leverage that led to a great deal of our success.
Â
So after selling that business, I thought, well, can we help other companies in this same way? Can we help scale this idea out? So and we started to see there'll be a trend toward, I think probably listeners, whether on LinkedIn or other places, have started to see this fractional word coming up more and more. So it's definitely on the rise. How do we help clients with that? Well, our goal is to have a vetted pool of talent, on the one hand, that we can help clients to leverage and help companies in need of that expertise.
Â
And also, I actually think that it's a bit different, all of us, for most of our professional careers, we've probably worked in full time roles with the expectation is that, yeah, everybody's around. I don't know whether it's nine to five, eight to six, whatever the hours are Monday to Friday. Of course, we're in the world of remote work now, but you're still available by slack or however, and so you're engaged in that kind of synchronous time period. And there's a lot of value in that, right, being able to go back and forth with your teammates in real time.
Â
We get these concerns from companies that, oh, wait, is that how this is going to work or like or how will I engage with someone fractionally? So what we realized is that we needed to help companies to understand you can get access to talent that you couldn't otherwise afford taking the fractional approach, not only because you're paying for only part of their time, but also the secret that we discovered was that a lot of really talented folks are willing to, in effect, discount their surplus hours.
Â
They're willing to, in exchange for having a sort of steady sideline, they're willing to kind of discount, and that makes it more affordable to work with some really tough folks. But we realized we need to help put some structure around how that kind of is organized on behalf of clients. Because, again, just making sure that the communications things and all of this stuff works smoothly, that's where we realized that we could add value in, sort of onboarding folks into this fractional world. And once they see it work, then like, oh, yeah, I need somebody to help with this.
Â
I need help somebody help with that. And while we started in the software engineering world, because that's sort of my background and capability I could help with. We broadened beyond that, and we've tinkered as far as helping with accountants, helping with, actually, Para-planners, in the financial sort of wealth management world and here's the furthest to field, telehealth, doctors. So even going that far.
Â
Richard Walker 5:47Â
Wow. Yeah, I was going to ask about that. Because I know a lot of entrepreneurs and leaders think of fractional CFO, even chief marketing officer, but I don't know how many think about their technology team being fractional, because traditionally, it's either you hire in house or you outsource. And even when you outsource, you're either doing a project or you're getting somebody allocated that they've chosen, et cetera. It's kind of out of your hands. But what you're talking about, I think, is a little bit more finessed, right? You're bringing this to the technical world. And I want to ask this question, actually, are the people that you're bringing in fractionally actually employed at Google or Facebook or some other company full time and they are actually giving you surplus hours? Or do they work full time for Fraction.
Â
Praveen Ghanta 6:32
Yeah. No, it's a great question. And firstly, I should note that we work solely with US talent, and the reason for that is, one of the things that we learned in our own experience doing this ourselves was that there's a lot of great folks here in the United States that are interested in this kind of work, and that's where we can really find senior level, experienced folks who have the capacity, but they also have the tacit knowledge in specific industries. And so some of the companies that we've been privileged to serve are in healthcare, in health tech.
Â
And so what we've found, and what they've told us, clients have told us, is that the difference in having someone who's actually experienced the American healthcare system, used it as a customer, it makes a lot of difference in the just tacit understanding of requirements, like, if you've never paid a copay, or versus dealt with all of the insurance sort of, like, is this approved? Is this in network, out of network? So in many of these sort of healthcare specific situations, and just to use one vertical industry, one example, there's a lot of value that comes with that. And so we found that that's a differentiation sort of with further afield. But in answer to your question, so of the talent pool that we work with, we actually found the strongest single source of talent for us nowadays is bootstrapping founders.
Â
Interestingly. So we're finding a lot of technical talent, they are starting their own companies, or they're interested in starting their own company, but they still need some cash flow pay the rent and whatnot, and so fractional work is one way for them to do that. So but that's the strongest in terms of, because when you think about it, they've got that founder hat on. They're thinking for themselves. They're thinking from a business perspective, what do they need to be doing for clients? So they're able to sort of step into some of those shoes? Well, as I'm helping, I don't know as I'm helping Rich to build Quik!. Well, what are these customers trying to do with these forms?
Â
Those kinds of questions, whereas oftentimes, whenever you deal with lots of folks who are great, let's say developers or great sort of workers, employees, they're not sort of putting themselves in that decision making space because they've never had to before. And so that's where we found, like a really good talent pool, as in founders. We do also work with some employees who are at larger companies, who maybe just have spare capacity just owing to their role. And we found that somewhere around the 10 year mark, particularly in technical roles, you'll find a fair number of really talented engineers, who they get stuck, and how they get stuck is they're maybe not the kinds of folks that want to be managers or manage other people.
Â
Maybe that's not part of their skill set or part of their interest. And so the avenues for further advancement in their career start to, some of those start to close off. It's like, okay, well, I'm not going to be manager to the team, so they kind of end up staying in maybe a senior engineering role and as often happens, like once you really got the reins on a particular role, it gets easier, relatively speaking, so you're still adding value for your company, but the tax that it is mentally for you is much lower. So they get bored at that point, and that's when we find folks like that, that some of them are interested actually in helping startups, precisely because it's interesting, right?
Â
And it scratches that itch. So, yeah, an example would be, there's a guy that we work with who is sort of a fractional CTO, not just a fractional C level, but he also does hands on work as well. And he actually sold a business. So he was a founder, actually, YC backed founder, and he managed to sell that startup to a larger company. Now he's gonna exit. He never needs to work again, but he is employed. He's got a great position as a director of engineering at a large company, large tech company, and again, to scratch that itch of working on cool projects, he's been working with us for the past six months.
Â
Richard Walker 10:29Â
Okay, now you're making me envious. I mean, I love my company, but I always love seeing startups and seeing what people are doing, and you're getting this window into all these different projects and opportunities, etc. I think that's amazing. And you actually described how we started, Quik!, so I think a lot of people know now my mom is my co-founder, and she and I were both technologists. We did consulting to pay the bills. I mean, the software product just wasn't paying the bills for years, and we just took on projects on the side to do it.
Â
So we were essentially doing what you're saying. It was project based for various companies. It was more consulting, but it makes a lot of sense. So that's awesome. And if you're hiring your company, you may be getting a lot more knowledge capital than just the software side of it. You're seeing big picture and other industries and other avenues that people have been exercising. I think that's fantastic. So I mean, you kind of said this, but what are some of the myths and misconceived notions people come to you with that they're trying to overcome, that you help them see differently, aside from what some of the things you said?
Â
Praveen Ghanta 11:34Â
Yeah, as I think this sort of fractional concept has become a little better known, there's still an assumption that a lot of is C level, or maybe sort of leadership level only, and that does work, and it's become more and more common that you'll have like a fractional CFO for a small company, because they don't need a full time CFO, or maybe even a fractional head of marketing, like a CMO, lots of roles in that vein. But what we found, and so this was in our background experience at HiddenLevers. So similar to the story you're telling Rich with Quik! and you were paying the bills by consulting.
Â
Well at HiddenLevers, I almost I pre-loaded that work. I was working fractionally before starting the company. While I just had my day job, I was also doing that, and I was sort of building up a cash reserve. Then whenever I started HiddenLevers, this is like 2010, or so, I knew that I had reserves. I did actually start fractionally as well, though I kept my day job and did it as sort of like a nights and weekend project. For the first nine months, we start to get, like, some inkling that, oh, maybe financial advisors are interested in this portfolio stress testing thing. So then we start at somewhere around month nine and month 10, I go all in full time, but I have this reserve that I built up from before.
Â
Once we got customer traction and we started getting revenue in the door, we couldn't afford to hire another full time engineer. So the fractional piece for us, the difference between hiring a C level person was we were trying to figure out, can we hire folks part time to help us build the software so actually write code, and that was the divergence for us, is that as we grew that team, quite a few of our senior engineers that were doing hands on work, they would work half time, so roughly 20 hours a week for us and that's one of the big misconceptions that we see in the market, and we're always educating around this, is that you can absolutely build a product with fractional engineers.
Â
Now, would I advocate that your team be entirely fractional? Of course not. We found in our own experience, with our own company, that the red line is somewhere between a third and half of headcount. And so on our product engineering team, for instance, it was five out of 12 folks. Five were fractional. So it 40 some odd percent. And that worked, but I don't think we could have gone too much higher than that. Obviously, you need folks around during the middle of the day for production support and all those
Â
Richard Walker 14:00Â
Well, you're also starting to answer one of my other questions I want to come up with, which is, are you losing out on knowledge capital and knowledge transfer if you have a fractional person, let's say this fractional person builds some component of your system, a web hook, an API, whatever, and then they go off to another project because you don't need them anymore. Do you lose that knowledge transfer?
Â
Praveen Ghanta 14:20Â
Right? Well. So the model that we saw working successfully for us was that our fractional talent, when we were engaged with them at Hidden levers. They actually stayed with us. So we had zero turnover in the fractional pool, and they stayed with us as long as any of our full time employees. So one person was with us for nine years. Another guy by the name Terrell, his case was funny. He and a couple of them actually made it across the acquisition. So that's a question we get sometimes, because, of course, we're working with startups. Is this a problem in M&A?
Â
And I can say from personal experience, no, because we had the four or 500 pages of M&A documents, and we had, like expensive lawyers reviewing everything and all of that past due diligence. And actually a couple of those folks went on to work for our acquirer, and one of them, as far as I know, is still there. And so yes, and so they're still contracting in this manner, fractionally and in helping support that, I guess it's not the Orion risk Intelligence Team and so, yeah, so that was able to continue. So that longevity and the sort of knowledge, human capital that you build up over time, it's absolutely possible to preserve that in this way as well.
Â
To be a little bit more precise about the working environment that, and this is the same that we do today with clients, our expectation is that these folks are attending two to three stand ups, like daily engineering stand ups per week. So they're part of that process, and they're part of the sprint review process, or whatever you know, kind of key meetings that a team has in place. They're there, so it shouldn't feel terrible different from working.
Â
Richard Walker 16:03Â
You want them to be integrated into the team, not just, hey, handle this ticket, handle this project. You want them to be part of the process and part of the overall team, right?
Â
Praveen Ghanta 16:13Â
Absolutely. Yeah. I mean, that's what we found is the most successful approach.
Â
Richard Walker 16:17Â
And do you focus on cultural fit as well.
Â
Praveen Ghanta 16:21Â
So that's an interesting question, because so one of the things that we did that we realized was very helpful, kind of, from a process perspective, was that the whole sort of interview process that a lot of companies use when hiring full time, that sometimes it's quite onerous on the client themselves, because it's like, oh, we need so and so to talk to this person. Now we need so and it's like, it's three rounds deeper or more, maybe it's five rounds deep, all of that. So in place of that, we view our job as being vetting the talents, capability, so engineering roles, all of whatever the tech stack may be, making sure that they're bringing that to the table, and we're also vetting them.
Â
So this was a key ad for us, one of the biggest challenges we've seen, because on occasion, hey, everybody's human so folks at different strengths and weaknesses, but a strength that we found that is mandatory for these kinds of roles is the ability to communicate on Slack or teams or any of these messaging platforms. Because that's how modern work gets done, right, but not everybody's great at that. And so we actually have taken to testing for that as well. So our goal is to bring so that's vetted, not just an oral communication, like we're doing here, but also the Slack team style communication, and then, of course, the core engineering work itself, so bring all those facets to the table. Now, does that mean that that person is going to be a perfect fit culture with any particular organization?
Â
No, not yet. That's a lot of the pieces, but that's not everything. And so there, rather than jump through a bunch of interviews, we do a risk free trial process, okay? And so, work with the person and see if they actually gel with kind of the way that you do things.
Â
Richard Walker 18:05Â
yeah. And then are your clients? Big clients, small clients, startups, other bootstrapped entrepreneurs, who's attracted to this most?
Â
Praveen Ghanta 18:14Â
Yeah, I think it's helpful to describe like, well, what's the ideal client, but what do we learn? And what have we learned about that? And so who can we help the most? What we found, in short, is that companies sort of between 10 and, say, 50 employees. So startups that are out of the garage, so to speak, or they've got revenue flowing, and they're trying to grow from there, which, interestingly, that lines up almost precisely with my own company and my own background experience with HiddenLevers and when we were able to take advantage of the talent.
Â
We get a lot of inbound requests from early stage founders who are pre revenue, pre product, and we've figured out a way to help them using low code solutions maybe to help build an MVP or something like that. But in practice, the my recommendation to anybody who ends up listening, who is a founder, who is thinking about that, is that, if you're the non-technical founder, that finding that person who can be the technical founder, I think it's still pretty important. As much as anything, you need a partner in crime, it's very difficult to do everything solo. And so we found that, yeah, it's hard for us, or for any vendor for that matter, to replace that.
Â
So we tend to fit in much better, if those pieces are in place, you know, then we're in a much better position to be helpful. On the large side, well, we do have some outlier clients that are a little bit larger. The vast majority are between 10 and 50. We joke that if you've got an HR department, you're probably too big for us.
Â
Richard Walker 19:46Â
I think that's a good one. I want to go back to this thing of having a technical copartner, co-founder, because I think what I want people to understand about that is when you don't have a technology background and you're working with a technologist to implement a technology you want, the expectations could be unrealistic or not understandable and not translate well from both sides. And having a technologist in the world who can actually speak geek and English, as I like to say, and translate the requirement is really the key here. I can't tell you've had this, I'm sure. Hey, I've got this great idea for an app. How much is it going to cost to build it and bring it to market? I'm like, nothing less than six figures. What do you mean? It's such a simple app? No, it's not. I'm sure you see that
Â
Praveen Ghanta 20:30Â
Yeah, so I literally had that conversation yesterday. So the funny thing is, and so what we do, what we recommend in those cases, when we talked to a founder that they don't have that technical person to work against is, you know, we're honest about it. We say, listen, in this case, and we learned this the hard way that, here's the thing that we learned the hard way. A lot of non-technical founders and the business founder, they'll come in, they've got an idea, and it may be a great idea, but there's this notion that, oh, we'll just build the app, let's say, and we'll hand it to you, and then it's done, right? And of course, we all know that the reality is that when you launch your MVP, that's the starting gun. That's not the finish line, it's literally the starting gun.
Â
And two things could happen from there, one your MVP, this, this app or product that you're putting out there, crickets actually, that often happens. And then you go through this whole process of trying to figure out who could be interested in this, and customer discovery and all of that. There's a lot of work to be done on that front. Now, it's possible that that doesn't succeed. And then how good or bad the app was probably was irrelevant, and it's sort of like neither here or there. But if you do succeed, which is the problem you want to have, what happens? Well, the customers start asking questions. They've got feature requests, there are bug reports, there's all of the things that come along, and now you're building a real tech company. And so, what we tell people is that actually, if you succeed, you're going to have more needs, not less.
Â
And that, I think, can be something that, sometimes for early stage founders, it's hard to wrap their head around. It's like, What? What? It's no, I just need the app. Well, you need the app, and then you need all of those requests and all of those bugs and all those things fixed on an ongoing basis, or the duration of your company, if you're thinking of yourselves as a software company.
Â
Richard Walker 22:22Â
Yeah, for sure, what's the adage it's an entrepreneur. But I think especially true for a tech company is somebody who jumps off a cliff and builds the parachute or the plane on the way down. And if you keep flying, then you got to keep building it and make it fly better. Man, it is such a journey. Okay, let's switch gears a little bit, because you, I'm sure love AI as much as I do. Being in the tech space, you guys are doing some pretty interesting things with AI, but I want to talk about how you're employing AI with your team members, and they're using AI with customers. So let's talk about that. How is that working? What are they doing? How is AI affecting software development?
Â
Praveen Ghanta 22:58Â
Yeah, for sure. So, and we really started to see this probably late last year, of course, all these tools come out, chatGPT and everything. And probably what early, beginning of 2023, is when it's a huge flash by then, and everybody's like, oh, wow, this is going to change everything. And of course, we're seeing continue to accelerate. But early on, a lot of developers were sort of playing around like, okay, can this thing really help us write code? And a lot of folks were sort of dismissive, like, yeah, you can use it for, like, to play around with and whatever.
Â
And I'll be honest, actually, even myself, in early experiments, I was like, yeah, seems modestly interesting. But what changed was that one of the fractional CTOs that we work with quite a bit guy, Jeremy, he was showing, he was demonstrating, firstly, with his work, an incredible degree of productivity, and was able to juggle more fractional clients than we would normally think possible, and do it successfully. And so I got to talking to him about like, his approach, and it caused me to change the way that I think about using AI. So I told myself, all right, I'm going to get into some code, and I'm going to force myself to make AI write it first every time. I'm not allowed to write the code myself. I'm only allowed to prompt, and that forced me to get a little bit better at asking, copilot, in this case, to what I wanted to write.
Â
So, write me some code that will do this and do it in this way, and use this library to do it and give it like, instead of giving it like the five word prompt or the one sentence prompt that sometimes we do know, let me make it a short paragraph at least, and really try to describe what I'm after. And what we realized is, like, you know what, in the same way that we would do if we were writing requirements for a more junior developer? Let's feed the tool as though it's a junior developer. So then all of a sudden, you see the quality of the code it's putting out. It's much higher, you know, is it high enough for prime time? No, it's not, but it's much closer. So, we realized that this pairing, and this is what Jeremy so his background is, he's actually sort of a CTO level person who still likes to be hands on.
Â
And so what we found was pairing generative AI with really senior talent, more senior talent than you would usually think of using in a role, it's actually a great match, because now they're able to generate code super-fast, right? Because they don't have to write it all themselves. And based on their experience, they're actually able to review that code. So it's like, pretend again, pretend the AI is the junior developer, so it's coming up with all of this, like, meh, to, like, mildly good code. Now you've got a CTO actually code, reviewing it in real time and beating it into shape, processes where we found, in aggregate, I would say, relative to a traditional engineering approach, about a 4x productivity gain.
Â
And I know that that sounds crazy, but we're finding that about half of that total gain, so the doubling of speed is the AI piece, and the other half is actually that you're using a CTOs time, and you're not asking a mid-career, or a developer with five years of experience? No, you're bringing a lot more experience to the table.
Â
Richard Walker 26:04Â
I think that is so impressive. And I think about my days programming, which are behind me for the most part, and some of the things I would do is I would use Microsoft Excel to automate building code, because maybe I had 40 property statements that need they're all formatted the same way, right? I don't want to type out 40 things over and over and over again, so I'd build a formula type in the 40 property names and boom, output the properties and paste them into my code. And AI can do that so fast for repetitive type stuff, and what you're describing also makes me think about how we write technical specs.
Â
If you write really good business requirements and really good functional requirements, and you take that and have GPT read it, and then say, now give me the framework of a technical spec, and then in detail, now tell me how to write this API. But you can't just say, give me an API. You have to say, here's what I want it to do. Here are the parameters I want the property. So it's the same thing with the code that you're talking about. It's not just, hey, give me a function that opens a file. I want these properties to go with it, and I want this functionality and these validations. So the prompting is really critical, and you're reinforcing something I keep saying. So I'm going to share this with my team, because I told them a few days ago. In fact, probably I'm going to give you credit.
Â
You've totally inspired me in our prior conversations. I don't want to hire anybody who doesn't have the skill set to use AI to help them code and the willingness. So, I've said this just yesterday, like when Microsoft Excel came out and was replacing lotus, 123, and when word was replacing Word Perfect. If you didn't learn those, you got left behind. And that's what this is right now, if you're not learning how to code better with AI as your assistant, you're gonna get left behind, because I'm gonna hire that person and not the one who doesn't have that skill.
Â
Praveen Ghanta 26:42Â
Yeah, no, absolutely. Yeah, and that's exactly what we're seeing. And so we adopt, or we adapted, our interview process so that we're now sort of generative AI first, even within the interview process, it's enabled me to sort of baseline assumption, I should expect that folks can accomplish more faster. And so I want to see that, whenever we're talking to candidates, we still see, I would say that of engineers that I speak with, probably more than half have not really, truly adopted it yet, like they're using these tools as sort of glorified autocomplete, which is a little helpful, but that's a tiny percentage of what it's capable of doing.
Â
And actually, yeah, just to build on your example, you were saying you're using Excel to build sort of a lot of repetitive code. Yeah, you can have these tools write a whole bunch of SQL statements for you, or other boilerplate kinds of things that you may not either you don't want to do it, or it's going to take you hours because it's so just this nuisance process and generate all that much more quickly, enabling, you'd actually spend your time thinking about the actual hard problems. Because the hard problems, by the way, the stuff that really needs doing, it's not there yet.
Â
Richard Walker 29:07Â
Yeah. I will say, I think one of the best uses of these large language models with coding is to figure out and do the research, figure out what you need to be doing. And so doing this recently, asking it questions like, which methodology is best practice in the industry, which would have a better speed which one would have lower storage requirements or cheaper storage requirements. And these are the tool sets I'm looking at, and it will tell you the pros and cons, or if you just say, here's a problem I want to solve, it'll come back and say, use this library. Use this methodology.
Â
And I think that saves so much research and development time in the process, because you're not trying to figure out, well, what component do I need to make this actually work? I just told you chatGPT knows, this stuff's fascinating, so I'm wondering if you're seeing, let me take another step back. A long time ago, there was an eBook that I read. I think it was called ship it now. You ever heard of that by chance? Yeah. Yeah. And so the concept for everyone to understand, I think the concept applies to anything you're designing, not just software.
Â
The concept was you need to boil down your design into very, very small, incremental steps. And so like in a screen that you're just designing, maybe you need a button. You'd say, design the button, and then the next one is, do the click of the button, and the next one is what the click does, etc. And you break it into these small parts, because then when you give it to a developer, they're responsible for that 10 or 15 or 30 minutes of work. And it's very clear what to do. To me, that's a lot of what the prompting is about. You have to be really, really clear on what you do. And the outcome is, if you're really clear and you understand it well, you have great quality code coming out and great solutions coming out. So I'm wondering if you help clients do that, or see it from that perspective, especially as you're putting AI into this world.
Praveen Ghanta 30:51Â
Yeah, we're actively looking at, of course, there's so many of these AI tools that have come out, and some of them are helpful, as we're discussing in the actual process of writing the code. Some of them are helpful in the design process, in the process of taking the designs and creating requirements. Now, each of those steps, let's say hypothetically, and I'm sure that folks may have heard about these demos of like, oh, they've automated the entire software development process, right? And the reality is, there's a bit of error. Like, I think on a good day, the output from a lot of these tools will be something like 80% on point, sometimes very occasionally, and I mentioned like SQL queries and things like sometimes it's 100% with stuff like that, but quite often, it'll be 80% quite often it'll be 50.
Â
At any rate, that's still helpful if you've got the right kind of senior talent, taking that as an input, cleaning it up before you get to the next stage. But you could imagine, if you were to prompt one of these design generation tools to generate, like, the design for these pages that have XYZ features on them for this new product I want to build so it designs some like, figma mocks for you. And then you said, okay, now take these big mocks and turn this into detailed technical requirements and populate Jira, or another project management tool. And so there are tools that'll do that. So that'll do that. Now take these JIRA requirements, and let's actually run this through code generation to actually generate something. Well, if you're introducing 20% error each step of the way, like a good case, right?
Â
Well, we can all do the compounding and see that the final product is going to be so far off the mark. And that's why, instead of just trying to chain all this stuff together, and I actually don't believe so, we could go way down the rabbit hole, like, how far is AI going to go. I'm more at least right now, from what I can see on the wavelength of it looks like we're limiting out, like we're kind of getting to out of it's a plateau, but it doesn't look like it's going to be able to do all of our jobs tomorrow. And so instead having the right, and in some cases, this is where the fractional approach comes back in, having the right fraction of the right person's time to audit and clean up each of those steps, well, you can still get a high level of productivity then without having, like, compounding errors where then it's total garbage out because it's like errors that are there as well in the way.
Â
Richard Walker 33:12Â
Yeah, look, I got to say something to anybody listening and watching this. I hope we didn't lose you with some of the technical talk and some of the words like figma mock you were referring to a design mock up. Sigma is the tool, etc, because the principles are the same. The principles are design things for each stage carefully, and you can use a chatGPT to help you along those stages, but you're breaking it up into nice components, whether that's a workflow in your operations or its actual software. So I think the principles are the same and that it applies universally. Praveen, I got to wrap up. I just realized how long we've been talking, and I've really enjoyed this. So I've got to ask you another question, but before we get there, what's the best way for people to find and connect with you?
Â
Praveen Ghanta 33:53Â
Yeah. So find us online hirefraction.com, or feel free to hit me on LinkedIn. Definitely pretty active over there, DM me or reach out.
Â
Richard Walker 34:02Â
Awesome. All right, so my last question, who has had the biggest impact on your leadership style and how you approach your role today?
Â
Praveen Ghanta 34:09Â
Yeah, so thinking about this, I actually have to give a shout out to my co-founder, Raj Udeshi, who is a co-founder of HiddenLevers, and so she was sort of my partner in crime for call it 11 years, and had a huge impact on as we built that company, in terms of leadership style and just sort of understanding. And the thing actually the lesson that I would that the biggest lesson I took away from that before I worked with him, I had a failed startup that failed because I was trying to do it too much by myself, and so I was hoping to find someone, whenever I started something again, that could complement my own weaknesses. And I definitely found that in him.
Â
So we had a complimentary set of skills, and we didn't often agree on things, at least not at the beginning of a debate about. Like, how should we make decision X for the company? What should the marketing strategy be? What should the business strategy be, etc, etc. And so there was lots of, I'm sure our employees will remember. It sounded like drag out fights, but actually it was very positive and constructive, because we would end up with a decision that was in between, and usually better than either idea that either of us started with. So I think that just learning about that, having that constructive relationship with folks that bring complimentary skills to the table, because you can't have a bunch of yes folks that are all the same, what they know.
Â
Richard Walker 34:11Â
I love that you don't get a polished stone without friction. So you don't get the best work without some friction in your company. Now, whether that's highly vocal or just, internal mechanisms, it's a cultural thing, I suppose, but I think that's awesome. I love hearing about that. And these people in your life make you better as a result. If you're willing to go through it with them, that's great. All right, I wanna give a huge thank you to Praveen Ghanta, the CEO and co-founder of Fraction, for being on this episode of The Customer Wins. Go check out Praveen website at hirefraction.com and don't forget to check out Quik! at quikforms.com, where we make processing forms easy. I hope you've enjoyed this discussion, will click the like button, share this with someone and subscribe to our channels for future episodes of The Customer Wins. Praveen thank you so much for joining me today.
Â
Praveen Ghanta 36:15Â
Thanks again, Rich.
Â
Outro 36:17Â
Thanks for listening to The Customer Wins podcast. We'll see you again next time, and be sure to click subscribe to get future episodes.
Comments