AIMinds #003 | Nisha Iyer, Co-Founder & CTO at CoNote

About this episode

Nisha Iyer is the Head of Analytics Services at Atlassian and Technical Advisor & Co-founder at CoNote (formerly CTO). Her journey started with a love for numbers and math, a passion she maintained despite her initial stint in Communications. Nisha's first career break came at Booz Allen, a prominent consulting firm, where she found herself managing large scale projects, some of which involved government websites. Her experience at Booz Allen sparked an interest in tech and led her to shift her career focus. Today, she encourages others to take control of their career paths and not hesitate in making a change if it aligns better with their aspirations.

Listen to the episode on Spotify, Apple Podcast, Podcast addicts, Castbox. You can also watch this episode on YouTube.

In this episode of AIMinds, we had the privilege of sitting down with Nisha Iyer. Nisha shared her journey into the tech industry, from a background in communications to diving headfirst into data science and AI, and eventually founding CoNote.

During this conversation, Nisha discussed the challenges she faced in her previous roles, particularly the disconnect between research, product development, and engineering teams. These challenges inspired her to create CoNote, a powerful insights engine that streamlines the research cycle, turning insights into action in a matter of minutes.

What sets CoNote apart is its focus on continuous research, a concept Nisha coined to bridge the gap between traditional research processes and the fast-paced world of software development. Nisha highlighted the importance of aligning user insights with product roadmaps and engineering meetings in a more succinct manner, ensuring that user feedback becomes an integral part of the development pipeline.

Using AI as a core component, CoNote seamlessly transcribes and synthesizes audio and video files, provides key themes and actionable insights, and simplifies the process of creating reports. The platform empowers teams to rapidly apply real-time user feedback to their product and feature decisions, enhancing customer empathy and driving impactful outcomes.

Nisha cautioned against falling into the trap of building features simply because AI can enable them. Instead, she emphasized the significance of identifying genuine pain points and developing solutions that prioritize addressing those needs, with AI serving as an enabler of efficiency and effectiveness.

Check out the full episode for an in-depth look at CoNote's journey and the role of AI in shaping the future of product development.

Fun Fact: Before CoNote, Nisha was part of an ed tech startup where she proposed the pivot from services to building a product, which led her to delve into product development and get into the entrepreneurial space.

Show Notes

00:00 Transition from arts to tech, pursued data.
04:06 New industry growth led to tech career.
07:09 Passionate about building product, disillusioned with data.
10:11 Frustration with feedback leads to building better products.
13:38 Continuous research is essential but often siloed.
18:49 CoNote born before Chat GPT in 2022.
20:33 AI could write prompt responses, but carefully.

More Quotes from Nisha

Transcript

Demetrios:

Okay. Welcome everyone, to the AI Minds podcast. This is a podcast where we explore the company of tomorrow built AI first. I'm your host, Demetrios. And in this episode, like all of the other episodes, it is brought. Brought to you by Deepgram, the number one speech to text and text to speech API on the Internet. Trusted by the world's top enterprises, conversational AI leaders, and startups like Spotify, Twilio, NASA, and Citibank. We are here with Nisha today, the CTO of Conote.

Demetrios:

Nisha, how are you doing?

Nisha Iyer:

Doing well, how are you?

Demetrios:

I am great and I'm excited to talk to you about what you've been building and how you've been getting into creating a product that is AI first and forward thinking and what that has enabled in your journey. But before we get into any of that, I would love to learn a little bit more about you and what got you into tech.

Nisha Iyer:

Yeah, that's a great story that I can tell. Yeah, I'm very excited to be here. Thanks for having me. So I'm CTO and co founder of CoNote. My journey into tech is definitely not a linear one, so I'm always an advocate to tell everyone, go for what you want to change careers if you need to. I did, and what brought me to tech was my love for numbers and math. My undergrad was communications and I was actually working at Booz Allen large consulting firm in this more communications project management type role. I was working on a large project that was based on a couple of government websites.

Nisha Iyer:

And so I worked closely with dev teams and I just started paying attention to what they were doing and was really intrigued more on about development cycle and I thought writing code was fun. And this is back in the early 2010, so like 2012, and I took a real passion to understanding more about analytics, Google Analytics, like why aren't we understanding where people drop off our website, that kind of thing. And at the same time, data science was kind of blowing up. I think back then it was like big data, right? There's always been some kind of buzword around ML data science. I think now it's AI. But I was really excited about diving deeper into a more analytical type of role. And because my background and my education had been so qualitative and more on the arts and humanities side, I at that time thought this is a good opportunity for me to go back and do my master's. And there was a new program, a data science programs had just started, so I went and did my master's at GW and that was, I would say my segue into tech was actually just getting to really work with those people while I was not in the development type role.

Nisha Iyer:

But I jumped in headfirst with the masters, and it was crazy. I had not done a mathematical or computer science based undergrad degree, so I was going back and having these very hard, high level computational optimization stochastic foundations just thrown into the fire. And I loved it. It was hard. It challenged me. And I felt like there was a lot of ways to apply data science to real world problems. And so I felt like I found my passion. It really interested me.

Nisha Iyer:

It felt like a new industry. I mean, it is a new industry and continuously is growing, as we've seen over the last decade. And I just felt like there was a lot of places where people had a lot of data and weren't really doing anything with it, including the job I was at that time. And so, yeah, I just started what really, I think made me have a strong journey and career path into tech, and I can go into that in a second, was the fact that I was not only just thinking about, what do I do once I get the data? I was really trying to think, what are the business problems? And then how do I take those and translate them to problems that I can address? If I ask for certain data, would I be able to address that problem using some type of descriptive, prescriptive, or predictive model? So I'll stop there. But that's kind of how I came into tech.

Demetrios:

Okay. I love that you almost had, like, two lives that you lived. You had your life in the industry, and then you said, you know what? I think that I can provide more value in the technical sphere. And I like that more. You dove into that. You went and got your masters, you came out, and you recognized, I imagine, from your previous life, the value of understanding the business needs around data before jumping into just, hey, let's do fun stuff with data.

Nisha Iyer:

Exactly. And I think that is, as I've grown in my career and moved into more leadership positions, what I see is lacking in a lot of data scientists and just in technical teams. Right. They really just get excited about the problem they can solve, but not necessarily getting all the business context, which then ends up with a really cool solution for a not so such a big problem. And then it's not really bought into because the problem is like, okay, we didn't need this complex solution. We could have just built some rules in Excel, and it would have worked. And so just, like, also understanding when there isn't, I truly believe most problems can be solved using data science and predictive modeling. Does everything need that? No.

Nisha Iyer:

When do you jump in and use these more complex techniques? And when do you say, let's do this in a more simpler way, and then let's save the more complex things for what actually needs to be run through a model or use speech to text transcription, that kind of thing.

Demetrios:

Yeah, that trade off is huge because you could probably get quite far with something that is more simple. And it is really like taking a long, hard look at when you need to spend that extra time to get that extra 5% to 10% that you're talking about. So then you did that. But I really appreciate the product that you're building now with Kono, and somehow you got into the product space. And can you give us a bit of background on what the inspiration was there?

Nisha Iyer:

Yeah, I love building product, and that has come about once again as my journey continued. So I'll just real quick go back to what happened after I got my master's. I quickly moved over to a data science heavy position. I was a data scientist at that same company, moved over to another corporation, discovery, and I was doing a lot of quote unquote data science, but it was mostly like metrics building dashboards, and I got frustrated. I was like, this isn't really data science, descriptive analytics, but not like, I'm not getting to do the cool stuff. Anyways, that led me to the startup I was at before CoNote, which was ed tech startup, and that ed tech startup built data science trainings for corporations using their data. So it was a customized type of training, but it was like, mostly this is like 2016. This was to get people to really grasp what you could do with data science, not just like, and so it spoke loudly to me.

Nisha Iyer:

So I was like, yes, I want to do that. So I was there for five years. I was the second hire. So I ended up leading the tech side of the company and built the entire tech team. So there was me and one other data scientist ended up. By the time I left, I had 50 people under me, engineering team, data science team, and designers. And while I was there, I proposed flipping the business model on its head. It was more like a services type of company.

Nisha Iyer:

I said we should be building product. And how did I get into product? I just started. I love to read articles and to understand what's going on and business models and how can you be most profitable. And what was SaaS. I actually work my full time jobs at Atlassian right now as head of analytics services founder life so that was one of the products I really liked. And I learned about their story and how they kind of built the initial SaaS model. And so as someone that was very invested in this other startup, I was like, hey, we should be built. Like, we could flip this on its head and use AI to automatically create trainings that are tailored to big clients.

Nisha Iyer:

My boss, and at that time also the founder, was like, yes, let's do it. And so I started my journey into product development. And the best way to learn is just to go headfirst into something where you don't have a lot of resources because you have to wear every hat, right? So as we were building product, I worked with my current co founder. He was on my team and he was a UX designer. Him and I ended up having to do a lot of the qualitative research, like the user research. We just had to wear a bunch of hats and we would do these great interviews, probably like 5 hours, like 1 hour each. And then we'd be like, oh, no, we have all this data. We can transcribe it, but we don't have the time to sit there and really find the themes and find the good pieces.

Nisha Iyer:

So what would happen is we'd just be like, oh, let's just take what we remember. And internally we're like, this is so bad. We could be building such a better product if we actually understood the feedback. But that either costs at this time, like days or weeks of our time, or we have to hire someone that's also cost ad. So that's how I got into building product was just like, I have a very entrepreneurial nature. I like to think of when I said that I think any problem can be solved with data science, I think that when I think about business problems, I'm also like, my mind's always spinning of like, oh, what could I build to do this? What could I build to do this? And when this happened, I talked to Cam, who's my co founder, and I was like, we could do like, I have the background. You're a designer. My other co founder, James, has taken products to market.

Nisha Iyer:

I was like, let's go. Let's do it. And so that's kind of how conote started, and that's also was my journey into product.

Demetrios:

Okay, so you mapped out a bit about what conote is, but you're kind of leaving us hanging. Can you break down what exactly the tool is and what it does?

Nisha Iyer:

Yeah. CoNote is insights to action. We are hyper focused on the research cycles that currently our beta is focused on user research. We know that this could be applied to multiple industries in various ways, and we do have other people using it even now, such as academia and marketing use cases. But the prime use case is focused on being able to speed up your research cycle. So really making research part of that DevOps process. Right now we have CI CD continuous integration, continuous development, but there's a missing piece which is top of the funnel, which is continuous research. And CoNote allows that to happen.

Nisha Iyer:

So what CoNote does is it allows you to upload your transcripts, audio or video files. This is where Deepgram comes in. They're immediately transcribed, and then we synthesize them using our AI engine. And our AI engine consists of multiple algorithms. It's not just one simple call to an LLM, although we do use llms. And after that synthesis, the user will get themes across the hours of interviews. So let's say I did 20 hours of interviews, I will not only get summaries of each of those interviews, I get the themes across all those interviews, which is like the hard thing that we found to capture. You also then get action items.

Nisha Iyer:

So you would get a list of action items that tell you what to do, how to take action on those insights you just received. And then you're able to create reports within the platform, which we also have realized as a pain point. So you're able to take the video clips, the pieces of specific text within whatever, 20 hours of transcripts, and create multiple pieces of those into a report that you can then send out to stakeholders. So it takes a process that usually takes days, weeks, as I said before, and turns it into minutes to a couple of hours. So it really allows you to really be agile with your research process as well.

Demetrios:

Break down this idea of the continuous research for me a little bit more because I like where you're going with that. I'm not sure that I fully understand it or grasp it yet, though.

Nisha Iyer:

So there's research ops. That's the thing. Continuous research is kind of like my thing coined it. But I think that what I've seen, like working at large corporations and at a startup that's not CoNote, is that researchers are siloed from developers, right? Like there is a user research team quite often, or product manager that's helping out with user research if it's a smaller company. However, because of the lengthy, tedious nature of user research and the amount of time it takes to distill insights and actually get something that's actionable, a research cycle might be six to eight weeks and dev cycles are two weeks. So by the time you get feedback on something that might have happened like last month and you've already built a new feature for it, you're like three features in, you're like, wait a second, this feedback is actually telling me I shouldn't have built the feature I released in two sprints ago. And instead, let's say I am a product manager working with a dev team and I do 5 hours of interviews. I schedule them like two days in a row, right? And so now I have these 5 hours of qualitative interviews and instead of saying, okay, user, okay team, now let's go translate these into insights.

Nisha Iyer:

And now dev team, we figure out the features based on our own drinking, our own koolaid, I would say I'm going to put this into CoNote, build a quick report, take it to my spread planning meeting, and be able to work with my director of engineering to figure out which features make most sense with the instant user research we just received about the latest information.

Demetrios:

So it feels like you've been able to take all of these needs that you had in your past lives and all of these times where you said, oh man, you know, what I need right now is something that I could see what people are gathering, what kind of data people are gathering out there, and I could implement it with what my roadmap looks like. And you just decided to go out and build that. And now CoNote takes those insights and is able to actively add them to the roadmaps, add them to the engineering meetings in a way that engineers, I presume, can digest in an easier manner. They are able to work with the product team in a more succinct manner. And the insights are also surfaced faster.

Nisha Iyer:

Yes, that is exactly what it is. I think because I think, like you said, from both your backgrounds, and I think it's because I've had the pain points of being on the dev side and then also on the more leadership product dev type role. And I think that from both those views, it really helped meld together the critical problem, which is that communication, like that loop between the research teams and the business and the dev teams, and honestly, there are engineers. When I worked at that tech startup, engineers wanted to understand the user problem. They don't want to be so siloed. But because of timelines and because of constraints and because we have to move fast and get something out, they get forced to just kind of make stuff up. I was just at a summit with my two co founders this last couple of days and we're sitting in a room thinking of really great things. I had to stop multiple times and be like, you guys, we're doing it.

Nisha Iyer:

We're sitting in a room making shit up when we really need to be drinking. At CoNote, we listen to what we're preaching, we practice what we preach. We do multiple rounds of user interviews multiple times and we synthesize it using CoNote. So that's another cool thing. But it's so easy to get in that zone of like, oh, yeah, no, I know what the users want and you don't get that empathy, right? And I want to be able to give that customer empathy to everyone, not just to a certain group within that entire ecosystem of building out a product.

Demetrios:

So there is another piece that I wanted to touch on real fast around how when you built CoNote, what I really like about it is that you said, there is a pain here and let's try and solve for this pain. But AI is only a piece of that. It's not like, hey, AI does all this cool stuff. Let's see if we can shoehorn it into some kind of a product. It was obvious that you said, you know what, I've had this pain. I've seen teams that I've worked with have this pain and now I'm going to try and build a product. And it just so happens that that product is leveraging the leaps and bounds that we've had over the past year with AI. It's not that.

Demetrios:

Oh, Chat GPT does summarization. Maybe I should create a product with that. Right? So how do you think about that type of thing as you're building product and not fall into the trap of, oh, well, AI can do this now let's create a feature like that.

Nisha Iyer:

I mean, that's a great question. I'm so glad you asked it because I want to start with CoNote was born before Chat GPT came out. So CoNote, we started talking about CoNote September 2022 and Chat GPT bull up the world in November, right? End of November. And I think that I just say that because 100% I agree. Maybe you aren't saying explicit statement, but I do not believe you find the solution and then you find the problem. To build a great product, I think you need to personally or know multiple people that are experiencing a huge pain point. And that's what the tech of it, the how was AI? But the what was the pain point? And that's what led me to want to build Kono was like, I experienced it, my two co founders had experienced it in different ways. We did a bunch of user research ourselves and talked to actual user researchers, product managers, academic people that also really like using CoNote.

Nisha Iyer:

And we got the same response, which is, yes, this process is so painful. And I also always like to say AI is a component of what we do. CoNote is not AI. CoNote is a insights engine that allows you go from insight to action in minutes. AI is a component that allows us to let users experience that. And I think that's how I differentiate that, because otherwise you're caught in this. We see so many startups, right? I see the kind of competitors, closest competitors to CoNote do something similar where they're just building all these templates that are obviously just prompts. They're like, AI could do this.

Nisha Iyer:

AI could do this. AI could do this. And it's cool at first, but anyone can write a prompt, right? That is not how you want to build something. And you highlight one sentence on one of our competitors platforms and it'll give you four paragraphs in response. That is not accurate. That's very obvious. So I think it's just, yes, being very careful, not falling into the trap and also just with my staunch data science mathematical background, being like, I'm not like AI. I mean, deep learning is AI, right? This has become a thing because of OpenAI and how great they've been able to market it to the people that didn't trust AI before.

Nisha Iyer:

So it's beneficial to us because now people do trust AI products. But I think that our goal is to allow intuitive usage delight users, get users from their outcome in a matter of minutes and allow them to get to the point they need to go, the next point they need to go to make impact. It's not to add really cool AI features.

Demetrios:

Well, Nisha, I am fascinated by the product that you are creating. I'm so excited to be able to follow along on the journey and I love the fact that you're part of the Deepgram startup program and we can help you out along your way. This has been an incredible talk. I thank you so much for coming on here and sharing your story with us. And with that, I think we will come to a close.

Nisha Iyer:

Awesome. Thank you, Demetrius, great having, great being had on the show.