Learn how Dr. Viswanathan and his team took their AI chatbot platform for recruiters from $0 to over $22,000 in MRR.
What’s your background and what are you working on?
My name is Vaisagh, but most people call me VT. I earned my PhD in AI & complex systems at Nanyang Technological University in Singapore. After that, I worked as a Post-Doctoral Research Fellow at TUMCREATE, where I ran simulations to study the impact of electro-mobility on tropical mega-cities.
Today, I’m the co-founder and CTO at impress.ai, an AI chatbot platform for recruiters. I lead our engineering teams in Singapore and India, as well as ensuring our product runs smoothly for our clients. Our conversational bots conduct competency-based structured interviews using techniques from Industrial Organizational Psychology, specifically situational judgment questions. They also autonomously interview, engage, and shortlist candidates at scale, 24/7, and actively fight human bias by hiding biasing information from human reviewers.
Recruiters at enterprise companies use our platform to save time, improve efficiency, and engage candidates during the recruitment process. We have 15 clients to date and an average monthly recurring revenue of US $22,000, a figure that’s growing around 20% month-over-month.
What motivated you to get started with your company?
In 2013, I started working with two friends on ideatory.co, a platform which helped enterprise companies hire data scientists. At the time, I had a steady full-time job so I was only working on this part-time, during nights and weekends. The business generated half a million in revenue, however, the product wasn’t scalable.
Then in 2016, one of the co-founders, Sudh, came back from a trip to Australia with the idea of impress.ai. He wanted us to leverage AI (since two of the co-founders have PhDs in AI) and create something unique in the recruitment space. This got all of us excited and I decided to quit my postdoc position to pursue impress.ai full time.
What went into building the initial product?
Initially Sudh created a chatbot using Chatfuel to test the market with our advisors and clients. While he was doing this, I started building the platform. Given my background, I knew how we could apply AI and machine learning to power up the platform.
We started using Telegram and Facebook before our own interface was ready. We also made use of API libraries to get the product working quickly. Over the next two months, I focused on creating the core internal, scalable architecture to get the first working version of the chatbot. That went live in September 2016 and two weeks later we had a basic functional “Recruiter Dashboard.”
The product went through countless iterations over the next few months. I spent the bulk of my time improving the conversation experience and then created an HTML widget so enterprise clients didn’t have to rely on the chat platforms. The product continued to evolve as we signed on more client accounts and their feature requests started coming in. For example, we created a full-fledged recruiter dashboard and built our own FAQ in early 2017. While we sold subscription plans initially and started seeing revenue within the first few months, it took almost one year before we had a product that was sales ready. We received initial support from Zeroth.ai, Asia’s first AI/ML accelerator, however revenue kept us afloat through tough times.
How have you attracted users and grown your company?
Our first customer, DBS Bank, signed on as a result of working with us in our previous business. From there, we were featured in the press, created a local meetup group, and started speaking at events which got the word out in Singapore. To expand internationally, we began setting up booths at conferences to test out new markets, like India, Hong Kong, and most recently, the United States. Also, we hired our first full-time sales executive and marketing specialist. Meeting people and doing live demos face-to-face has helped generate the most interest. Creating a new website where visitors can request demos and using LinkedIn as a tool for lead generation has been most valuable in terms of attracting new prospects for sales.
What’s your business model, and how have you grown your revenue?
We have multiple tiers for different levels of enterprise needs. My advice to other entrepreneurs who are looking to grow revenue is to factor in how much time and effort it takes your team to deliver results for clients. This goes beyond the product itself, and includes things like integrations and training. Also, it’s important to study the market and understand how your competitors are pricing their platforms.
What are your goals for the future?
In terms of the product, we plan to launch a self-service version ready where recruiters can create their own chatbots without our delivery team getting involved. We also plan on launching conversational bots in other languages, including Mandarin. In the long run, there are quite a few ideas in the works that will make the whole process much simpler for candidates as well as recruiters. I’m really excited about this but can’t really reveal much about this yet.
In terms of my team and the company, I’m working towards scaling and maximizing the team’s potential. Right now, I have the exact kind of team that I always wanted — talented, young, passionate, and itching to prove themselves. I want to create a culture that they are proud to work for our team and an environment where they achieve their full potential. I’m still learning how to do this. Apart from having a great product, it’s also about helping individuals realise and articulate their vision and set them on a path to achieving that. When they realise they can achieve what they want in their career by working with us, we are all winners.
What are the biggest challenges you’ve faced and obstacles you’ve overcome? If you had to start over, what would you do differently?
Hiring the right people for our team hasn’t been easy. Initially, I focused on getting talented people and didn’t focus too much on whether they would be a good cultural fit. It was at the back of my mind, but I didn’t take this into consideration during the application process. Later, I overcompensated in the other direction. I hired a candidate with excellent work ethic but very little experience. I assumed we could train him, but working a startup with a lot of pressure, we just didn’t have the bandwidth to provide the training and attention needed, and it became a drain on our resources. As time went along, we created a company vision and that has made hiring processes easier.
Another challenge has come with the increased number of candidates interacting with our chatbots; we needed a way to scale infrastructure but didn’t have enough time or resources to hire an experienced DevOps person for our team. As a result, I was forced to learn DevOps fairly quickly to keep the platform running smoothly.
I’m also transitioning from building the platform to trusting a team to enhance the platform. I’ve started creating processes, which is a delicate balance of prioritising tasks and assigning them to the right people. In theory, it sounds easy to follow a textbook template but in reality, most people will stop following this if they find it too difficult. So, when making decisions that impact how our team works, I’ve started thinking about how useful as well as how sustainable the processes are. I’ve learned a lot from each of these challenges, and I wouldn’t have done anything differently.
Have you found anything particularly helpful or advantageous?
I like reading, and that has helped me learn about products and technology. Currently, I’m reading Principles by Ray Dalio, it’s a brilliant book about running successful organizations. Two of my favourite books about design are Don’t Make me Think by Steve Krug and Design of Everyday Things by Donald Norman. Books that I’d recommend every programmer to read include Programming Pearls, Code Complete, and Structure and Interpretation of Computer Programs. Another book I swear by is Dale Carnegie’s classic, How to Win Friends and Influence People, because it’s packed with wisdom.
I’m at an advantage because of my talented co-founders, Sudhanshu (Sudh) Ahuja and Amrith Dhananjayan. We trust each other to manage our respective parts of the business. For example, I can focus on the technology aspect of building a kick-ass product because I know Sudh will bring in the sales and investment, while Amrith will manage clients. Over the years, I’ve realised that having this sort of synergy in skills has helped us get further.
Also, I’m pretty good at explaining complex concepts in understandable terms. This is especially valuable when it comes to a product like impress.ai, where a significant amount of time and effort is spent educating others, both internally and externally about the product, the technology, and the possibilities that arise from it. Also, my ability to do this well has helped me receive valuable advice from stakeholders.
What’s your advice for entrepreneurs who are just starting out?
Be ruthless about prioritization. There will always be a hundred different things you need to execute and this can easily overwhelm you. But as long as you’re able to follow a workflow and differentiate between high vs low priority tasks, fire-fighting in times of crisis will be the exception rather than the norm.
Another key thing is reaching out to people for help. That initial ask requires a level of humility and strength that doesn’t come naturally to everyone. In the early days of working on your startup, you need to let yourself be vulnerable and admit that you don’t know everything there is to run a business or build a product from scratch.
Where can we go to learn more?
Our website is a good place to start.
You can also follow impress.ai on social media, we’re on LinkedIn, Twitter, and Facebook.
If you have any questions or feedback, leave them in the comments section below or email [email protected].