Ever struggled with navigating complex software systems? Imagine a tool that simplifies this process with the help of AI, making your tasks easier and more intuitive.
That’s exactly what TruvaAI offers! This innovative tool transforms user onboarding and retention with intelligent AI agents that perform tasks within apps, reducing the hassle and complexity.
Welcome to an exclusive edition of our Interview series on AllAboutAI.com, where we introduce various stakeholders and developers in the AI industry to share their stories, challenges, and insights. Today, we had the pleasure of speaking with Gaurav Aggarwal, co-founder of TruvaAI. This conversation highlights Gaurav’s journey from software engineering to AI innovation, the strategic development of Truva, and the future of AI in digital adoption platforms.
Here’s the full interview, available on our YouTube channel!
Transforming Digital Adoption: In Conversation with Gaurav Aggarwal, Co-Founder of TruvaAI
Let’s talk about the fascinating insights Gaurav shared with us about his journey, the inception of TruvaAI, and the transformative potential of AI in digital adoption platforms.
Which career are you in, and how did it help you become an AI stakeholder?
I’ve been involved in machine learning and AI for a long time. Originally from India, I have a background in software engineering. My journey in machine learning began unexpectedly when I joined a team at Microsoft in India around 2011-2012. They wanted me to train classifiers, and at the time, I had no idea what a classifier was.
It wasn’t that I knew a lot about AI initially. It was more about being in the right team at the right time and learning on the go. I was fortunate to have colleagues, mentors, and managers who nudged me and helped me understand. One thing led to another, and I developed an interest in AI.
From there, I pursued my career at LinkedIn and took machine learning classes at Stanford. That’s how I got into AI. I wouldn’t call myself an AI researcher; I like applying AI to real-world problems. It’s both a passion and a profession for me. For example, where I live, the street in front of my home has too many cars. I thought of using a Raspberry Pi to count the vehicles using a model. It’s just fun, but that’s how I feel.
What inspired you to create Truva, and how did you develop the idea?
The idea for Truva came from my co-founder, Anuja. She was very excited about the LLM space overall. Her thinking was that using AWS daily can be overwhelming due to its complex interface. Often, you want to perform a simple task within this vast system. This is similar to HR software like Workday and Rippling; as a founder, I have no interest in becoming an expert in these tools. I want to get my job done, such as changing payroll or adding employee leave, and then get back to where I want to focus.
Initially, I had no idea there was a space called the digital adoption platform or that there were tools around it. Our initial thinking led us to realize this space existed, with large companies solving it for in-house employees and others for external users. With the advent of agents, we saw an opportunity to make the experience more personalized and efficient.
When logging into software, I don’t want to go through a lengthy product tour with numerous steps. I want to know which steps are essential and how to return the tour if I skip it. The process can be painful, especially for a small startup without dedicated account support.
These are the problems that could be addressed more cohesively. Thus, the idea of an AI agent account manager was born, aiming to solve these issues as a package.
How does Truva differentiate itself from other AI onboarding and customer retention platforms?
We see ourselves as…
It’s a very competitive market. Customer onboarding and workflows are highly competitive areas that have been around for close to 10 years. Truva will help you in 3 ways;
- First, if the task is simple, we’ll do it for you. For example, if you go to a CRM and ask for insights about a particular deal or calculate a forecasting score, we can do that using your logic and rules. You don’t need to review all the content and synthesize it yourself.
- Second, if we can’t do the task directly, we can trigger personalized tools for you. Often, with product tours, once you skip them, it’s hard to bring them back, causing FOMO or anxiety. But with our agent in the app, you can ask, “Show me how to change my personalization settings,” and it will guide you through it. If there’s an API, we can make the changes for you directly.
- Third, we handle all this complexity and provide a turnkey, easy-to-use solution. We are your solution provider.
From the software provider’s perspective, setting up these tools traditionally takes months of planning and maintenance. They often use solutions like Google Dialogflow, which require a team to manage all possible user interactions. This process is time-consuming and requires in-house expertise. That’s where we come in.
Can you share some success stories or case studies where Truva significantly impacted a client’s business?
We are currently working on case studies. We started as a plugin and launched it last month. While we don’t have official integrations with significant sources like HubSpot, we are available as an app on the marketplace and a Chrome plugin. We have also been distributing our app to customers through a partner strategy. We collaborate with partners who implement HubSpot, Pipe Drive, and similar systems for businesses to get our app into the hands of end customers.
These partnerships have seen up to a 25% improvement in implementation time and task completion.
CRMs often have issues like data duplication, where multiple entries for the same person are created. For example, if Gaurav is already in your CRM, another entry for the same Gaurav might be made, resulting in duplicate data. People need nudges to get the data right. If the data in a CRM is incorrect, the leadership relies on inaccurate information.
What were the biggest challenges you faced while developing Truva?
It’s a work in progress right now. We just started in the last couple of months, and there are a couple of challenges.
First, TruvaAI is a technically challenging product because people don’t want to give full context. Even I wouldn’t do it. For example, if I’m on a deal in a CRM, I might say, “Hey, create a follow-up and send the meeting invite to the stakeholders.” But follow up to what? What is a follow-up? How do you create a follow-up? Who are the stakeholders? There’s a lot of information to infer from the context.
We need to determine from the page you are on whether it is a deal page, which deals it is, who the stakeholders are, and how to create the follow-up. Each step requires a specific agent. This complexity needs to be handled in a human-friendly and timely manner, which is a significant challenge for us.
Second, some customers and leads have concerns about AI. They appreciate the technology but are worried about AI, likening it to a Terminator scenario. They fear how much they can trust it, especially with the current news and discussions around AI.
We are addressing these apprehensions and determining how to best cater to those concerns. It’s understandable, given the stories we often hear about AI.
How does Truva’s AI technology work to enhance customer experience and retention?
Depending on the integration, you can inject a JavaScript snippet to show it in your app or use a Chrome plugin. It’s a small widget that appears, allowing you to chat, search, study problems, and come up with answers or trigger workflows. It can either appear as a side panel if it’s a Chrome extension or as a widget within the app.
What are some key features of Truva that you are particularly proud of?
I’m proud of the time I spent getting started with Truva. As I mentioned, if you use Dialogflow today, it’s a bit of an eyesore. You have to figure out all the possible ways to start a conversation and what needs to be done. It’s like creating a flowchart with all possible combinations. Not everyone thinks in that structured, logical manner.
It’s a very engineering-driven process, and if you’re a growth marketer or a CSM wanting to make your customers happy, you’re not interested in this entire flowchart setup. It takes a long time, and you need engineering resources.
Our agents handle most of these complex tasks for us. About 90% of the work is on us. For you, it’s more about approval or enabling/disabling certain functionalities. If you’re uncomfortable with agents doing certain things, you can say, “Hey, don’t do it for me. Don’t change anything in this setting.”
This addresses the Terminator theory concerns, ensuring you have control without us being aware of those actions.
Can you discuss any upcoming features or updates for Truva?
We are adding event tracking where the system is self-learning. Right now, the system is self-learning, but the learning cycle is quite long—about weeks—where we track all user feedback.
This comes from my background in search engines. For example;
If you search for restaurants in Palo Alto and click on the second result, you will not like the first result because you must have seen it before.
This is subconscious feedback. Similarly, with our agent, what you do after a conversation tells us whether you are satisfied with the conversation and the action tools. With sufficient data, our models can improve.
We are working to improve this process much faster, aiming for almost daily updates. We’re not there yet, but aim for faster than weekly updates. This developmental stage offers endless possibilities. You don’t always know exactly what the user wants, which can also be new to us. Having this kind of responsiveness is fantastic.
How do you see the future of AI in customer service?
A critical aspect of customer experience or customer service is that…
For example, Figma is an exciting tool. I used Figma for something a while back, and I probably made some clicks indicating I was confused. Figma asked me, “Hey, do you want to do this?” It was like having a guest Figma mate. I thought, “This is pretty cool.” I wish all tools had this kind of experience. That’s the experience people are gravitating towards.
However, you must be careful not to bug users too much. If you constantly ask, “Do you mean this? Do you mean this?” it becomes frustrating. You have to get it right to create the “aha” moment. You want to read the minds of consumers, and in the future, we might have AI that can read consumers’ minds to some extent.
What AI and tech industry trends are you most excited about now?
I’m a big fan of the agent approach. In the day, reinforcement learning was how you made agents and was a math-driven approach. I barely used agents, even in my 10 years in large companies and various hobby projects. But with the advent of these LLM agents, it’s much easier to create agents with specific optimization functions. Imagine a small team of folks with different roles assigned to them, able to cater to various problems.
These agents are still naive, more like interns, but they will get smarter as progress is made. That’s where we’ll see a lot of action and improvement on the AI side for the foreseeable future. The next decade or so is for the agent.
How do you stay updated with the latest technological advancements and trends?
If I had all the time in the world, I would love to read research papers.
But I do read some interesting blogs and follow a few newsletters. One of my favourites is the Alpha Signal newsletter. I also follow people like Yann from Facebook and Andrew Ng from DeepLearning.ai. I research when something interests me, but only about one in twenty times. I don’t have enough time to review all the research reports.
What leadership qualities do you believe are essential for managing a successful startup?
All startups are the same unless you’re doing a foundation model. Foundation model: if you’re building a foundation model, that’s a very capital-intensive business. It’s basically like medical research in marketing; you know how medical research startups are. But for everyone doing applied AI, like fine-tuning existing models, we are almost the same as any other startup.
We aim to work with our first set of customers, proving our value in the pilot phase. We don’t want to scale prematurely. We need to prove our value, address all their product needs, and overcome apprehensions about AI, such as the “Terminator” fear.
How do you foster a culture of innovation and creativity within your team?
Our culture fosters creativity and innovation. We have a very flat hierarchy with no managers. Everyone has goals and extreme autonomy. For example, we have an open mandate:
If there is any SaaS tool you think will make you more productive and under a specific dollar amount, let me know, and we’ll get it.
For me, productivity is of utmost importance.
Our team is a mix of people from startups and large companies. We have folks from Facebook, Google, Amazon, and startups like Flexport. We take the best practices from all these experiences to create a unique and influential culture.
What has been the most rewarding moment in your entrepreneurial journey so far?
The most rewarding moments are when I’m on a call with a customer who says they’ve been having issues with something. For example, I told you about the Dial-A-Flow situation and figuring things out. Someone whose job it is to build that will understand the pain and effort it takes. If you’re a people manager who hasn’t done it yourself, you might not appreciate the complexity or the time needed.
So, whenever I’m talking to an actual stakeholder who is hands-on, that’s the moment. It’s a validation of the problem we’re solving.
If you weren’t leading Truva, what alternative career path might you have pursued?
I would still be involved in technology, working somewhere or in a specific space. I’m very fascinated with Artificial Intelligence, so I would be on a data infrastructure team or in applied AI, helping in some capacity.
Do you believe AI will replace human jobs in customer service?
In 50 years, but not in the next 5 to 10 years, if you’re creative, you’ll find your place using AI alongside human problem-solving skills.
I read an exciting line somewhere: “AI will not replace human jobs, but AI will replace human jobs that are not using AI,” which I think is true. If you think about it, there’s a hierarchy of knowledge-based jobs. For instance, there are cloud jobs, and then at the top, there are researchers, strategic thinkers, and business leaders who need data.
Many jobs, like data entry, were significant back in the day but are now largely automated. Humans will need to move up this hierarchy, becoming more like knowledge workers and problem solvers rather than performing standard data entry tasks.
People worry that generative AI will replace musicians and graphic designers. It will not completely replace them. AI might do a better design job than I can, as I’m not a graphic designer, but I can still use tools like Photoshop and Illustrator.
Is there anything else you’d like to share with our followers about Truva or your journey?
There’s one thing I knew but have been validating: I thought there were two types of entrepreneurs when I started the startup. Type one includes those already in the space, like someone doing a PhD or working in a large company, who worked on a project that got open-sourced and turned into a startup. They had no intention of being entrepreneurs; it just happened because they were brilliant, and the project took off.
The second type includes those who want to do a startup for various reasons. They try many different things. I thought both types could exist independently, but I’ve realized that an actual Type 2 entrepreneur eventually has to become Type 1.
You have to learn and become almost a world-renowned expert in the space you’re playing in.
What’s next for you and Truva? Are there any exciting projects or goals on the horizon?
We just started our journey and are marching towards our first set of case studies. We are converting our pilot projects, proving our value, and working from there. We are working hard and hope to have some case studies out soon.
How can our readers stay connected with you and follow Truva’s journey?
If you can add value to your software, please book a demo at Truva.ai. This will set up a demo with me or one of my co-founders. If you want to follow me, I’m active on LinkedIn. Drop me a message—I’m always happy to help if possible. That’s all from my side.
What advice would you give young entrepreneurs looking to start their ventures, particularly in AI?
To young entrepreneurs coming up and wanting to start their ventures, especially in the AI industry, I encourage you to keep doing what you’re doing. The community is an excellent resource, and many intelligent people are willing to help. If you reach out, you might be surprised at how willing people are to assist; they’re busy.
Even a cold message can change your life, so contact people. The worst case is they won’t reply, and that’s okay.
Key Takeaways
Our conversation with Gaurav Aggarwal provided valuable insights into the innovative journey of Truva, an AI-driven tool designed to simplify user onboarding and retention. By integrating AI agents within applications, Truva transforms how users interact with complex software, making tasks more intuitive and efficient.
From Gaurav’s background in machine learning to the inception of Truva, it’s clear that the tool is poised to enhance digital adoption platforms significantly. As AI evolves, solutions like Truva will improve user experiences and productivity.
For the latest updates and to see Truva in action, follow their journey on their website and social media platforms.
