I'm Ajitesh Abhishek, Co-Founder & CEO at Tough Tongue AI & Archie AI. Ask me anything!

Thanks for the question.

For me, the MBA was, in some ways, a ticket to work at a Silicon Valley company. I had a choice between doing a Masters in Computer Science or an MBA. I felt my background was more aligned with an MBA because of my experience in product management and my love for building products. I wanted to stay on that career path. (It's another story that I've been doing a lot more coding in last one year because of Archie AI and Tough Tongue).

That was my main reason for pursuing an MBA. Another reason was that I believe an MBA is what you make of it. I used my MBA to improve my teamwork skills—learning how to lead a diverse, international team through team projects and real-world challenges. Also, I also used my MBA time to do courses in machine learning and data engineering in Northwestern University. It's typically not a good idea to do this, but I clearly knew what I wanted to get out of it, and this turned out to be helpful when I joined Google Cloud, which was a deeply technical PM role.

On your second question about whether I would recommend an M7 MBA: I think it's a tricky landscape now, especially when you consider brand value and all that. But you do need to hone skills like leading a team in various situations, presenting your ideas effectively, doing customer research, asking the right questions, and marketing your product well. Some people pick these skills up naturally, while others learn through practice. An MBA provides a great platform to engage with people who have diverse experiences in these areas, and your professors often have real industry experience. It offers a good two-year environment where you can reflect, work on these skills, and try new projects. As long as you don't get too theoretical, and you work on side projects and learn actively rather than relying solely on the brand name, I think it can still be a good decision to go for an M7 MBA.

Thanks for the questions.

There was an internal model at Google (I think it was called "Meena"), released about a year before ChatGPT. It attracted a lot of discussion because it could say offensive things and wasn’t very guarded. But when you interacted with it—asking it to do or explain things—you could see clear magic. It was never released because of safety concerns and the risk of bad publicity.

Then ChatGPT came, and everything shifted. I have to give credit to Sundar, who pushed the idea of being bold and responsible at the same time. Google took some risky bets, even if that meant things could have gone differently. For example, serving AI queries increased our costs, and fast launches might have led to negative PR. But we made it through that time.

When the ChatGPT moment came in Dec '23, we were way behind ChatGPT, and it was a scary time for all of us. But there was also a strong belief in our technical strength, and everyone knew that the game had just started—it wasn’t over yet. I clearly remember a Google Cloud meeting where our CEO rallied us with a war cry that we would fight it out, and that was the start of something big.

Once all our objectives, OKRs, and team focus were set, it felt like we were in a fast, well-oiled machine, racing ahead. Then came the vision of multimodal models, and it was clear they were coming. Google was one of the first to put a stake in the ground, saying we would build a multimodal agent. That turned out to be a great goal. Just look at the video and audio parts of our models—they’re among the best in the world.

Another key point was the context length. While Anthropic raised it first, Google came in with a one-million token context window, and internally we even had about a 10-million token window. That was a clear advantage. On the technical side, especially in the cloud, we saw that some things ran a lot better on Google because of our design choices—like TPUs and how they work with storage. Some vendors even preferred Google Cloud for building these models.

It was really interesting to be in the middle of that race and experience everything firsthand. Overall, my thinking has changed a lot. I now believe that the model layer will be dominated by just a few players with tight margins, leaving little room for many others. On the other hand, the application layer will be the big winner, much like what we saw with the cloud and the internet, opening up many new opportunities. Also, we’re still very early in creating a great UX experience in the AI space. A few years from now, we might look back and think this was a magical new way to do things—something we haven’t fully seen in many app spaces yet.

Those are my thoughts—feel free to reach out if you want to talk more.

I was the first person to join the team after three co-founders, who I knew from college days. I got founders equity in Fitso—a good percentage of it. When I left, I had to give up a part of it, but I didn't sell the rest of it even in the subsequent rounds. During the pandemic, we weren't sure how things would work out. VCs continued to reach out to me to get a bit of outside-in info on Fitso and I was helpful. I got an exit when Zomato acquired us, and while I can't share complete details in a public forum, it was a good exit for me and all the early employees. All done in good faith!

Hi, good to be speaking with you. I decided to move back to India because it gave me more chances to build a startup. In the US, I was on an H1B, and if I wasn't working, I only had about three months to build something. I know some people can do it, and many do, but I felt that time was too short for what I wanted to achieve.

Another way to do a startup is O1 visa, which many try for, doesn't allow a spouse to work. That made it a complete no-go for me. And finally, I liked being close to my family here. I knew I would miss out on the support system I had in the US—the network of founders, the hackathon ecosystem, and the larger market opportunities. But after weighing the pros and cons, I was very keen on starting a startup, and I felt that being in India was the right move at that time. We'll see how things pan out. I'm optimistic about my decision.

We are not funded, but we're part of the Google Cloud Accelerator and NVIDIA Inception program. This gives us access to cloud credits and contacts in Google Cloud for support when needed. Yeah, it's expensive to build an AI product, and some costs are paid out of our own pockets. We're trying to be very frugal and have made choices to reduce costs on our model—for example, limiting how long users can engage in a single context session. Right now, we want to focus on building the product and gaining traction. Fundraising can be a huge distraction and exhausting exercise at times, so delaying it to time when it's absolutely necessary.

I love Google. It's an amazing place and was a dream for me too. I still miss going to the canteen, having access to baristas, the buffet of food, and working with such exciting people. But then I started to wonder, "Do I want to do this my whole life?" I felt it was the right time to do something I always wanted—build something on my own, make decisions that match my principles and values, and ride the AI wave, which is a once in a lifetime opportunity. It felt like a ticking clock, even though there will always be another chance. It just felt right for me.

I still love Google. If you're still there and want to invite me to the campus, I’d love to visit sometime. Thank you.

It wasn't any precise moment for me. I gradually reached that conclusion. I was working on a Gemini product at Google and saw huge demand and excitement from enterprises who often lag in adopting new technologies and trends. I felt that if I missed this cycle, I might not get another chance for a long time. That feeling kept growing. I was also working on side projects, and it didn’t feel like a risk—it felt like if I didn’t leave, I’d lose a real opportunity.

At first, most people didn't like the name and gave me a lot of negative feedback, saying it was a terrible name even though the product was good. But over time, as people started using it, friends, family, and users remembered it and truly got that Tough Tongue is about difficult conversations. At one point, I was changing it to Scenario Labs (https://www.scenariolabsai.com/) but every user I asked told me they like Tough Tongue now. It might sound a bit odd, but that uniqueness makes it memorable. That's why I've kept it, though I'm still open to ideas.

Consumer offering at this time. We could have distribution via coaches and business.

Yes, I code and my co-founder/CTO also codes. In the early days, we were head down—discussing, designing, and coding the product. As it matured, a lot of my time shifted to customer discussions, sales, and marketing.

On the journey from -1 to 1, everyone has their own approach. My approach is to start with some conviction, build something, and then see what sticks in the market, rather than doing tons of research upfront. I think that defeats the purpose of a startup, which is to experiment and see what works. You can bring deep expertise, but it's not like only experts win.