The Future of Low Code AI w/Jonathon Reilly
“Everybody approaches a problem or solution space from their frame of reference. It’s really all you have is your frame of reference.
If you wanna learn, you’re gonna need to get exposure to people with different frames of reference, right? You have to be open to being wrong...
Acknowledge that you’re wrong before anyone else points it out. That’s a career accelerator for you.”
“if your company doesn’t adopt machine learning and everyone else does, you’re going to be operating at a massive disadvantage. And sooner or later that company will go out of business and you’ll, if you’re working in that company, going to have to get a job at a company that has adopted machine learning. And, just like that, that’s the adoption curve reality.”
Great things in business are never done by only one person. They're done by a team of people. Low Code AI has the potential to empower more people to contribute to the development of groundbreaking products and services, expanding the possibilities for innovation.
Today’s guest — Jonathon O’Reilly, Co-Founder & COO at Akkio
The Future of Work Area — Low Code, Machine Learning, Artificial Intelligence
Listen to the full episode :
In this episode, we're joined by Jonathan Reilly, the Co-Founder, and COO of Akkio, a Low Code AI start-up that democratizes access to ML models and allows integration into business processes backed by Bain Capital. With a background in operations, product, and marketing at Sonos and Markforged, Jonathan brings incredibly valuable insights to the table with his vast experience in scaling products and operations, as well as experiencing companies’ scale from 0 to 1 and 100.
Jonathon and I had recently caught up during the D3 (Digital, Data, Design Institute) Catalyst event at Harvard Business School. He had a compelling discussion to educate the audience on how best to leverage Low Code AI, democratizing access to ML, and the potential impact of generative technology like OpenAI’s GPT models and Stable Diffusion on the business processes.
Jon’s personal story of going into entrepreneurship and embarking on a journey within the AI and product space is an inspiring one. We discussed challenges in maintaining the company culture while scaling the business and shared compelling insights into MLOps integration with RPA (robotic process automation) software, highlighting the potential for these two technologies to work together to streamline and automate business processes.
Throughout our conversation, Jonathan touched on the ethical considerations surrounding the use of technology within AI and the responsibility of technology companies to ensure their products are being used for good. He also provided valuable advice on how to approach building high-performance teams in a beneficial culture for a technology startup.
If you're interested in entrepreneurship, AI, and the technology space, this is an episode you won't want to miss. Tune in to hear from an experienced leader in the industry!
Some things we discussed:
How can product managers and engineers work together to identify market gaps and the business value of features?
What are some effective strategies for improving collaboration and communication between product managers and engineers?
Can you share your story of how you got into entrepreneurship and your experiences at Sonos and Markforged?
What led you to start Akkio, and can you talk about the early days of the company?
What motivates you to continue building products, and how does Low Code AI fit into that vision?
How can generative AI algorithms like OpenAI be integrated into Low Code AI platforms like Akkio?
What is the one thing that excited you about AI? What are some ethical considerations surrounding the use of technology within AI you run into?
Transcription
Alp Uguray, Creator & Host: Welcome to the Masters of Automation podcast episode. Today we have Jonathan O'Reilly with us. Jonathan, welcome. It's a pleasure to have you with us.
Jonathon Reilly: Thanks. It's great to be here, Alp.
Alp Uguray, Creator & Host: So Jonathan is the co-founder and COO at Akkio. Akkio is a low-code AI application platform. Before Akkio, Jonathan worked at Sony as an engineer, and then that led his path to be a product manager at Sonos where he's seen the startup in early days as well as its scale. And then, later he joined Markforged where he became a product leader and led marketing. And Markforged is a 3D printing company for both industrial goods as well as beyond. So now he is running a startup called Akkio that also my previous high school classmate, Ekin, works at. So it's a pleasure to have you today and have the opportunity to hear your story. And you have a very diverse background. You were an engineer, you were in product, you're now in operations, you led marketing, you've seen startups grow and scale as well as you worked at large companies Sony. And now you guys are kicking off your own startup and you already, I think, locked Series A as well. So what is your story ? what led you to Sony and then later to Markforged? where did it all start?
Jonathon Reilly: Yeah, well, I guess it kind of goes back to undergraduate. And I grew up in Montana, interestingly, and my father was a professor of marketing at Montana State University in Bozeman. And I was kind of disenfranchised with the education system in high school. I didn't really think it was teaching me much. And I decided in my head as a rebellious teenager that I was not going to go to school. I didn't want to go to universities. I could tell it would be the same thing that was happening in high school. But my dad, he told me I couldn't leave the house until I applied to a couple of schools and then even picked out a couple of schools for me to apply to, one of which was actually Gonzaga University, where I ended up going. And at the time in high school, the one thing that was really interesting to me was speech and debate. I was on the policy debate team and we were ranked number one in the state of Montana. So we were pretty good. We went to nationals. And Gonzaga had a debate program. And so I got into that school despite not trying very hard in high school. And I was going to do pre-law originally. And I get into the pre-law classes and they're really much too easy and it's kind of the same story. I decided to go to Gonzaga because I realized that college was going to be a party. I'm , this is going to be great. I can just have fun for four years. And once I had that realization, I'm I'm definitely going to school. But so I got into the pre-law classes and they were not engaging. And so fortunately, I was taking some of the advanced math classes and physics classes because I'd been in some AP classes in high school and that turned into electrical engineering. And so how did I end up at Sony? Well, one day in my senior year, I sat down on monster.com back in the day and I applied to every electrical engineering job in California. And I wrote up my application and my cover page to say what I wanted to do, which was design work. And a couple of companies got back to me. And when I reviewed the ones that got back to me, it looked I'd written my application specifically to their job. But that was just about the fit. I really just used the shotgun approach. And so I got hired and moved to San Diego to start doing design engineering on video processing circuitry for CRT televisions back in the day at Sony. And I did that for seven years. And it was really exciting, lots of learning. But eventually, it kind of got a little bit slow. It's each year, you're going to do the same thing over and over again, design the next TV and they ship a new model of TV every year. And sooner or later, you get a little bit bored. But I'm working at the time with a lot of product managers who are making the decisions about the features of the television. And I had a lot of questions about how those decisions were being made as an engineer. And I think this is a very, very traditional engineer product manager interaction where product management asks for some feature and engineering is , that feature doesn't make sense for one reason or another. And product management is , do it anyways. And so as an engineer, you're , OK, and you do it. And then later, they reverse. I have a good anecdote. There was a time where they asked us to put in this extra audio processing chip. And it would push the bill of materials over the limit that the bill of materials could be at by about $3. And we're , if we put this chip in, the bill of materials is going to be $3 over. And they're , do it. And I was , OK, so we put it in. And then the next roll up from the factory comes back and it's $3 over. And they come down and they're , we got to take $3 out of the bill of materials. I'm , really? It's , you got any ideas? I'm , well, there's this audio chip. And sure enough, I was , take it back out. And so anyways, one of the product managers who's actually pretty good that I was working with, he took a new role. And they asked him who he thought could replace him. And he actually suggested that I might make an interesting candidate because I was always questioning their decisions and trying to understand the logic behind it. And so I moved into product management, my first product management role at Sony. And I spent about a year and a half working on stuff there. And it was actually really beneficial to be able to take your engineering knowledge of the technology space and then meet that with market understanding of opportunity and value propositions. And I found that incredibly fascinating because you not only had to think of what would be good to do and why it would be good to do it, but make sure that it was actually possible to do within the constrained decision-making environment. So that actually, that part of it ended up being really exciting for me. The part that I d a little bit less was Sony is a really big company. And it's hard to have a really big impact as a young person coming up at a really big company. I mean, you can, but you're kind of always injected in these HR systems that are designed to govern your growth slowly and stuff that. But anyways, so I'm working in product management. And this guy, Phil Abram, who is the head of North America marketing for Sony TV, ended up leaving and taking a job at Sonos back when they were very small. And they were hiring product managers. And he suggested that I apply. And I met the team there. I'd never really heard of Sonos before. I was , I don't know what this small company is. I'm working at the number one, at the time, Sony was number one in television. I'm at the number one television company here. But then I met the team. And everyone I met was some of the smartest people that I'd met. And so I made the decision then, I'm going to work with these people. I'm going to learn from them. It's a startup. May not work out, but I'll have an opportunity to learn quite a bit. And as it turns out, it did work out. And in the working out, I grew my responsibility and career and ended up leading product management on a bunch of the products that they built, many of them still shipping in some iteration today. And watched it grow from, I think it was something $60, $70 million in revenue when I joined to almost a billion in revenue when I left. So it's a full scale up.
Alp Uguray, Creator & Host: You've seen the full scale up story.
Jonathon Reilly: Pretty big scale up opportunity there. And it's interesting because you hit these on the way up. When you're growing, you always hit these stall points, I think. And I've seen this happen multiple times now where you have an idea and you bring it forward. And maybe that's a new product category or a new feature that you bring to market. And it grows really fast for a while. But once you lap that growth, you're back to growing slowly again. And you need something else to keep growing. There's this constant hunger for growth if you're not growing or dying. But it was just a really fun run. And towards the end of it, when it was really big, I really missed those early startup days where it was , your life is on the line if you don't make the right decisions. Things are going to go poorly. It's just very motivating to wake up every day and realize that you might not survive. The company might not make it. And the responsibility for getting it across the finish line, for keeping it alive, is on you. And for me, at least, that ends up being super motivating because I just want to make it work and do whatever it takes. And so I actually had my second child, my daughter. And on paternity leave, I spent some time thinking. And I'm , I think I'd to jump into an earlier stage startup. And that led me to Mark Forged, a totally different industry. But I asked that product manager I'd worked with at Sonos, who he thought in the Boston area was interesting for me to speak with. And he made an introduction to Greg Mark, who is the CEO of Mark Forged. And I met the team there, Abe and Ekin and everybody, actually people that are co-founders in Akkio today. And it was the same basic experience I had when I first joined Sonos, which is I thought this is a group of really, really smart people that I could learn a lot from. And early on at Mark Forged, I asked, before I took the role of leading product there, to speak to some of the customers. And it was clear they had emerging product market fit. , they 3D print continuous strands of carbon fiber wound inside of nylon matrices. And these parts replace tooling and fixtures on manufacturing lines. This is the core value proposition at the time. Of course, they print lots of things now. And the people who were maintaining these manufacturing lines were paying 10 times as much for machined aluminum. And it was taking 10 times as long to get those parts in. So it was an order of magnitude improvement in their workflows. And they just loved the printers. So there was something really already there at its core. And we built it and launched new printers and scaled the company up. And it was a wild, fun ride, just as stressful and exciting as I imagined it would be. And towards the end, I ended up taking on some more responsibility in marketing when we had some issues with our lead pipeline and deal flow. And I realized the thing my dad taught at Montana State University was marketing. And so I'm an engineer. I'm never going to be in marketing. And so here I find myself actually running marketing. I'm , I'm not so sure I want to do this. But it turns out, actually, a lot of the optimization in a B2B marketing workflow is around data pipelines and efficiency and doing some math on figuring out where your best dollar spent to closed one deal sits and how to invest against the market of opportunities of advertising to maximize your qualified lead funnel. And there's always these tricky incentive structures between marketing and sales. In this case, marketing was incentivized to make marketing qualified leads. I think that's pretty typical. But the definition of those leads had been established a long time ago. And it's just kind of apparent that if you could use machine learning to match the firmographics and the title and the behaviors that a lead that came into the system had taken to your already closed one or closed loss business, that would be a much more effective way of deciding who to reach out to first, how to engage, how aggressively to engage. It basically make the entire pipeline run more efficiently. So this founding story of Akkio is we started looking for solutions on the market that would let us do things that and create smart robotic process automation workflows based on natural language input from users who filled a form. And we couldn't really find anything that let us build it. There were some solutions we tried on the market that had somebody else build it for us. They didn't work very well. And we realized that they were also application specific. And one of the problems is application specific solutions tend to be least common denominator solutions. The people in application specific ML solution has a right to live because you're an expert in the application, not necessarily an expert in ML. And there's always this tension between the business group and the person delivering a machine learning model of understanding what the objectives are and what the data actually says. And this even exists internal to companies. So the spark of the idea there was it would be really nice if there was a platform that allowed people to take an arbitrary data table for any application you might have and model the outcome of interest, understand the patterns that are driving that outcome, and then in a couple of clicks use it for real time decision making. And that idea is the core idea of Akkio. And of course, we've built lots of stuff since that original idea. But the main thesis was that ease of use and enablement would make it so that a lot of people who today do traditional data analysis in Excel or Tableau or Power BI could leverage machine learning in their processes and extract more value from their data faster. And that would be meaningfully impactful for a ton of businesses. So assuming that any data driven process in any business will eventually run on machine learning, the market's incredibly massive. And that's what we've been up to building Akkio. So we founded it and raised a friends and family round and then seed round from Bain Capital and made some progress on the A round. We haven't announced anything yet, but made some progress there too.
Alp Uguray, Creator & Host:That's a very impressive story. And I think throughout your journey, you've seen how a company can not only scale, but can maintain the culture that it has. And I think over time, that is one of the tough parts to maintain.
Jonathon Reilly: Culture is really hard, particularly if you're scaling. I mean, some of the times in scaling, you double your head count in one year. And if you think about that, half the people in the company came in from other cultures and the other half the people there, maybe half of them have only been there for a year by the time this influx happens. It takes a lot of considered thought. I think the leadership of Sonos did it quite well. We worked on it really hard at Markforged too. And it stretches from onboarding and orientation to setting examples and leading through examples every time you can. And we spent a lot of time thinking about that at Akkio too, making sure that we're building the right type of culture. And the things we focus on for what it's worth are, really we want a culture of ownership. I think your most successful companies and the only way that you can really effectively scale is if you can empower people to have agency. This concept of hiring agents, I talk about a lot because hiring someone who will just do what they're told and execute and do that fast and efficiently, that's good, but it's not sufficient, right? having someone who will think through the problem space, take ownership, be able to make decisions or come back and say, this is how we should solve this problem and why, and defend that point of view, that's what you're looking for, I think. And sometimes culture gets very political and it gets very hierarchical and you have , I can't talk to the person who's my manager's manager with an idea because my manager would get upset that I skipped leveled them or something or went over their heads. I frequently say, if I ever hear that happening with anyone reporting in my infrastructure, whoever was in the middle doesn't need to be there anymore, they can just go, right? Because I think the only way you get that agency is a collective ownership, shared incentive structure with stock, and then really, really empowering people to make decisions and take ownership. But again, very difficult to do. And it actually gets harder in my experience when companies get bigger and growth slows down because then the stock component of the compensation becomes less impactful. And when everyone's fighting over some salary dollars, kind of quickly becomes a little bit of a , a political game to try and get the promotion to SVP or something and that's your win condition. So, there's this idea I think, and we talked about it, Mark forged of two dreams in China, it's the dream of your personal success and then the dream of the country success. I think it's actually pretty similar at a business and you want both dreams to be alive and well if your culture is gonna be tight.
Alp Uguray, Creator & Host: Yeah, definitely. And I think there's the transparency and the visibility of the goals of the team and the company and the individual when they align well and top off the motivation that there will be more success later on if you continue to accomplish more. I think becomes a very sweet spot in those deals that which is very interesting because I think there are a lot of companies who struggle to scale. Maybe after you reach 1000 people then it's no longer a small friends, family and then I think group to community then to become a country at that point.
Jonathon Reilly: I think in my experience, the tipping point's been around 300 people, 200 to 300 people somewhere in there it tips over. , we'll see if we get that big at Akkio which I hope we do. We'll see how we can hold that off. And still hopefully have a culture of agency.
Alp Uguray, Creator & Host: Yeah, I'm sure you guys will accomplish that. I think the, well, one of the interesting aspects was you've seen the how Sony was, how Sonos was, how MarkForged was and seeing both as an engineer's perspective, a product manager's perspective, as a marketer's perspective. I think there are a lot of lessons learned for Akkio that maybe the company could have done better in the past or maybe they could have strategized certain things better. So what are some things that you learned over time that when you guys were starting Akkio that you all sat together and you're , okay, this is the three things that we are doing for sure.
Jonathon Reilly: I don't know that we were that explicit. The, I mean, obviously we were gonna build a really great product experience. Really what I think the trick is, is to be really clear with your strategy and your decision-making frameworks and to set those up early because people need, the way that a company operates and its sort of thesis on the market opportunity is built on some some principle thoughts or some constructed frameworks. And when someone comes new into a company, they don't know, and they certainly haven't spent a lot of time thinking about what the frameworks for decision-making are. And so practically the only way to get everybody rowing in the same direction as things scale is to be really clear and explicit around here is how we prioritize the product roadmap, for example. Here's the priority framework, and here is the strategy. And a good strategy is an umbrella. , it's there's things that are inside of the umbrella that are on strategy, and then it excludes a bunch of things you could do that would be off strategy. So for example, at the beginning, and by the way, strategies change and can be refreshed periodically, but at the beginning at Sonos, for example it was fill everyone's home with music. We were very focused on the home. So nobody came in and would suggest some out of home experience for audio that we would focus on concerts, or I don't know what else, there's a long list of audio experiences you have outside of your home, and probably lots of interesting opportunities for a sound company to go invest in those areas. But it was sort of a focusing lever, so to speak. And I think just getting those sort of , here's what we do, here's what we don't, and why things detailed out, and then telling them to everybody, and then telling them to everybody pretty frequently, and summarizing each bit of it as in as pithy a manner as you possibly can so that everyone can sort of hear it, remember it, and then sort of remember what unpacks underneath it, because no one's gonna remember three paragraphs describing what a strategy actually means, but they'll remember the name of the strategy, and then that'll parse into , oh yeah, this is why we're doing what we're doing. So strategy and decision-making frameworks, I think are incredibly necessary, and they're also very aligning, because as you make them, everyone either has to agree or disagree, and if you're not on the same page, then you gotta dig deeper and find where your fundamental assumptions differ, right? Because if you share a set of fundamental assumptions, you should be able to construct a shared view of what to do. And so almost always there's some , there's some fundamental assumption difference, and it's good to get that dragged out into the light and look at it from multiple directions. And I think you always also have to have room for debate. you have to be open to being wrong. , you can't be right all the time. And when you're wrong be the first person to stand up and admit it is my other tip. , acknowledge that you're wrong before anyone else points it out. That's a career accelerator for you.
Alp Uguray, Creator & Host: That's a very good point. There's the, in the Kim Scott's book, Radical Candor, there's a meeting structure that she introduced at Google and Apple. And then she says that when you're in two opposing directions and then there's an argument to be made, the first meeting is each party protect their own idea and then try to make an argument to win. Whatever it may be, whether that's about product strategy or maybe a new feature edition or maybe competitive intelligence, to blend into a new market. And then the next meeting, she's saying, okay, everyone now reverses the role. And now you need to defend his idea or her idea and vice versa. And then she's saying that , she finds that experience really interesting because then you have to study why the other person thinks differently.
Jonathon Reilly: What's really interesting is , if we go all the way back to the early days is , if we go all the way back, to policy debate, when I was in high school, that's what you do. you are either affirmative or negative, and you have to take either side of the topic each time you debate. And so , I think out of all high school sports I might recommend to someone, I think that would sort of be at the top of my list if you would call it a sport, maybe a mental sport. But it sort of forces that exact thing, which is you have to think through a problem from both sides of view. And then you have to understand all of the arguments really well. And then you can come to a conclusion of what you think is right and defend it and be passionate about your direction because the important thing is you've sort of divorced your sort of personal worth in being right from the direction. And you're really trying to just objectively look at the issue as best you can, right? , and I find if, and this is hard for anyone, this is hard for me, if you think you're right about something, you kind of don't want to hear opposing evidence it's , I think that's a little bit of human nature, honestly. And so you actually have to work at being right and being open to opposing evidence and thinking it through and then deciding , oh yeah, that convinced me or yeah, I see that point, but , here's why I don't think it's right.
Alp Uguray, Creator & Host:I think we are biased by what we know by nature, right, because we defend that idea. And so it's very interesting because then that is a fundamental reason why an innovation can happen. Because then if both parties started to debate on a topic and then they argue whether it may be, and then that leads to new technologies to be born or look at different perspectives. Then I would to tie the discussion to the event at Harvard, which was really nice, which was really cool event. And it was very interesting because at the event, the main thing was low-code AI. We discussed a little bit of OpenAI products and generative technologies in addition to RPA. And what I found very interesting was how the ideas of students, the tech leaders, the professors and maybe product managers, engineers currently at Google and whatnot, talk differently. So , as you were speaking, what were some of the highlights from the session for you?
Jonathon Reilly: Well, I think, actually, I think the most interesting part of that presentation was the Q&A session at the end where we talked about ethics and responsibility in machine learning. , I think to your point, everybody approaches a problem or solution space from their frame of reference. It's really all you have is your frame of reference. And so if you wanna learn, you're gonna need to get exposure to people with different frames of reference, right? And you have to be kind of open to that, but it doesn't mean that everyone's point of view makes a ton of sense or that they've really thought things through, but it is good to get exposure to really smart people's differing frames of reference because then you'll learn, right? And so I think one of the reasons I product management and really love working in product management is that I can learn.
What you're trying to accomplish has to make sense to the sales team, has to make sense to marketing, has to make sense to engineering about building it, has to make sense to the investors. customer support has to believe that , it's going to be a good experience and that and that it's a meaningful value add , you have to , you kind of have to work through all of those reference frames in order to , have a good position on what to do. And I think , the difference between good and bad product management really boils down to your ability to really explain to everybody, why, right? , and not everyone's always going to agree with you. I mean, there's been there's always an engineer who thinks you're wrong. , regardless, I could just guarantee that that's always going to be the case. There will always be a really smart engineer who thinks you're a complete idiot. And, and the challenge is going to be explaining or convincing them to look at it from a slightly different reference frame than they're operating in. And, and being able to do that. Being able to , tell the story and explain why and it does help to have some technical background. So your asks are anchored in reality of possible implementation. That's, that's kind of the trick to a good product manager. And , I've been on the other side as an engineer getting asked terrible, dumb things that don't make any sense. Some of them even implementable, but I just didn't believe the market , we can build this thing and it would take a lot of time and effort. I just really don't buy that this thing is going to sell. So it feels I'm going to do a bunch of work for no reason. And that that as an engineer, you hate that. Right. And so, so sometimes it's , Hey, look here's a way we can approach this problem that makes it tractable. But sometimes it's , here's why this market is set up for this product to really succeed and how it's going to work in the market. And, and a lot of times if you're doing something innovative, those types of things don't exist. It's, it's really kind of the fuzzy front end. You're dealing with a progressively loading JPEG problem, right? And you kind of got to say the picture is starting to resolve and it kind of looks this. I think the, an interesting part to that is the, especially in the field of local AI, ML, maybe automation, workflow management, there are different stakeholders and each of the stakeholder need to hear the story from their own way. the business, you said, maybe cares more about time efficiency or improved customer experience. Whereas the engineer cares more about how they can rebuild on a machine learning algorithm easier and without going through all the tests or retraining for hours. And then similarly for the end user, then experience a better product that actually works and functions. So I think in communicating the local, the entire AI and ML fields, there's very nuances and different value propositions to everyone. And, and where do you see, and then obviously this varies by audience, but the people getting challenged the most to adapt their own skillset, because they also need to acquire a new skillset themselves and also adapt to new way of doing the work and, and, and I think we've seen that often that there are processes in large enterprises that they're done by for 25, 30 years and somebody has been doing that. The government is typically the worst at these things, right? They built a system 50 years ago and are still using them. It's kind of an interesting question and I think the pace of change is accelerating actually because technology is going to happen faster and faster. And we've, we've clearly hit this interesting inflection point in AI and machine learning right now. And, and the the way you used to do things is going to change sort of for everybody in some capacity and that, that could be as, as, as basic as trying to learn about a topic using the internet or as complex as standing up an optimized workflow for your sales pipeline. And , I think, I think a lot of people get a little bit uncomfortable with change. , it's, it's sort of a, it's sort of a core reaction that, that even I have change can be the only constant changes, but change can be a little bit scary. But I found that, that the strategy of leaning into uncertainty and change is where you learn and grow the most as a person regardless of if that's at work or at home or whatever it is to be really true. And now I almost seek it out because it's where you get the most growth. And so, everyone from an analyst to used to do a sort of exploratory data analysis and graph key trends and then try and show what's driving them to , a data pipelining engineer. They're going to have new tools at their disposal that are going to make them more effective and more efficient. And , I think, I think one of the key skills to differentiate yourself is going to be a willingness to adopt early and to learn and to lean in to these types of changes. Because if you wait around, eventually the change will get forced on you from competitive purposes, right? , you're, if your company doesn't adopt machine learning and everyone else does, you're going to be operating at a massive disadvantage. And sooner or later that company will go out of business and you'll, if you're working in that company, you're going to have to get a job at a company that has adopted machine learning. Right. And, and that that, that's the adoption curve reality. There's always a group of people leaned in more on the front end of things. And there's always a group of people who have been doing something the same way for a really long time and don't want to change , the late adopters or the late majority as they might be.
I think the pace of change is accelerating and the time you'll be able to wait it out is getting shorter. , as, as a, so I think everyone's got to be flexible and lean in as much as they can
Alp Uguray, Creator & Host During my interview with Joy Mountford, she mentioned that sometimes she sees some engineers focus on mastering the tools. So, for example, somebody is really great at Python, certain package or be really great at a certain vendors product. And then she was saying, if that product or that programming language one day disappears, then need to learn the next thing. So understanding the principles of how each work give you the transferability of that knowledge. I would to tie this to you Akkio. I think it's a good generative tech future where you'll be able to come in, and you have the data, and then you can type in and prepare the data.
Jonathon Reilly: It's a good example where, we, what we do there, in case people aren't aware because, because it's new is we, we take a natural language request , and that could be anything reformatting dates or doing math or calculating things. And we translate it using a large language model into SQL to apply to your dataset. And then, we show you a preview of the transformation that would happen to the data. We subsequently take the query that we generate and feed that back into a large language model and ask it to describe what's going to happen verbosely. Because one of the key things is putting the human in the loop of this process because it turns out while people might be pretty good at specifying exactly what they want with code, they're a little bit less good at specifying exactly what they want with natural language. So, showing the interpretation of the ask helps them think through the level of specificity that they need to have or the corner cases. It's funny because it's the classic product manager engineering interaction right laid bare and a user interface. , I asked for a feature and the engineers come back and they're , you didn't think about these three corner cases. What do we do here? And then I go, Oh, I got to, I got to specify what we do there. Same thing. But, but yeah, when, one of the interesting things about lower no code or using natural language to instruct a computer is these are all just layers of abstraction to machine code, right? even, even Python compiles into machine code to run on, run on a computer. So this is just another layer of abstraction and with it comes some complexity around the specificity of event handling or edge case concerns. But also with it comes the ability for a lot more people to use a mechanism they already use every day language to start to instruct systems to do what they want. And I, that'll be incredibly powerful, right? that'll change everything because more people can do it with less barrier to entry. It just it saves massive amounts of time and effort for the people doing it the old way. But a bunch of new people who couldn't figure out how to do some, I mean, I can't tell you, even, even myself sometimes in Excel, I find myself Googling how to do V lookups and, and how to make a V look properly.
