Shreyas Nair Left Sequoia to Build an AI Startup From Scratch

Shreyas Nair spent years evaluating startups from the investor’s side of the table. As a Vice President at Sequoia India & Southeast Asia (now Peak XV), he backed growth-stage companies across SaaS, AI, and consumer internet. Before that, he cut his teeth on strategy work at McKinsey and earned an MBA at Harvard Business School.

In late 2023, he decided to switch seats. He co-founded Wordsworth AI, a company building AI agents that help e-commerce brands create personalized, high-converting landing pages and web experiences without heavy engineering lift.

The company is still early. It’s a 10-person remote team that is navigating one of the trickiest questions facing AI startups right now: how to balance a revenue-generating services model with the push toward a scalable product.

In this interview, Shreyas talks about how he acquired his first customers through founder-led outreach and free pilots, why he thinks the line between services and software is harder to draw than most founders expect, and what it took to build conviction as a first-time founder after years in finance and investing.

How much revenue and profit does the business generate?

We have done roughly $200k in annual revenue, and we have intentionally reinvested most of it back into product and team rather than optimizing for short-term profit.

Tell us about yourself and your business.

I’m Shreyas Nair, Co-Founder of Wordsworth AI. We’re building AI agents that help modern digital businesses create personalized, high-converting web experiences without needing heavy engineering bandwidth. In simple terms, we help companies move from generic websites to intelligent, conversion-optimized surfaces that adapt to their customers. At Wordsworth, I lead product, working closely across engineering, design, and customer success teams split between the U.S. and India.

Before starting Wordsworth, I was a Vice President at Sequoia India & Southeast Asia (now Peak XV), where I focused on growth-stage investments across SaaS, AI, and consumer internet. I started my career at McKinsey, working on strategy and operations for large industrial companies, and later completed my MBA at Harvard Business School.

At some point, I realized I wanted to get my hands dirty and build from the ground up. That shift led me to start Wordsworth, where we initially focused on e-commerce brands and now work with some of the fastest-growing digital businesses in the U.S. We were also selected for the Sequoia Capital India Spark program early on. What excites me most is this shift from static software to systems that can think, adapt, and continuously improve outcomes. And Wordsworth is our way of building toward that future.

How does your business make money?

Today, our revenue comes primarily through an AI services-led model. We work closely with e-commerce brands to design and launch intelligent landing pages and web experiences, and we charge for that work as a high-value AI-enabled service. This has allowed us to work on building AI agents that can actually help our workflow and solve the real customer pain point.

What was your inspiration for starting the business?

The original inspiration came from watching how much friction exists between campaign intent and execution. A marketer can have a sharp idea, a strong offer, and real urgency, but turning that into a live, high-converting web experience still takes too many people and too much time. By the time the page goes live, the moment has often passed, or the creative edge has been diluted.

We felt there had to be a better way. AI was getting good enough to help with ideation, structure, copy, and workflow orchestration, but most people were still using it in disconnected ways. We wanted to build something that sat much closer to the actual work: understanding the campaign, generating the right page structure, helping teams ship faster, and improving performance continuously. That gap between possibility and execution is really what pushed us to start Wordsworth AI.

How and when did you launch the business?

We started Wordsworth AI in late 2023. In the beginning, it was not some big, polished launch with a perfect product and a grand go-to-market plan. It was much more founder-like than that: a lot of customer conversations, rough prototypes, fast iteration, and manually doing things that we hoped software could eventually make repeatable.

Our launch was essentially customer-led. We got in front of brands, showed them what we believed the workflow could look like, and then earned trust by helping them solve real campaign problems. That approach gave us two things early on: revenue and very honest feedback. In hindsight, I think that was the right way to do it because it kept us grounded in actual pain instead of building in a vacuum.

Tell us about your team.

We’re a lean team of 10 today, including the founding team and a small group of engineers, customer success, and growth. The company is remote-first, with people working across geographies, which has pushed us to be intentional about communication, ownership, and velocity. Because we’re small, nobody gets to hide inside a narrow job description. That kind of operating style is demanding, but for an early-stage company, it’s been a real advantage.

How are you funded? 

We raised a small angel SAFE round in early 2024 of $800K. We have since used that capital and supported the business through cashflows. 

How did you acquire your first customers?

Our first customers came through direct founder-led outreach, relationships, and a lot of persistence. In the early days, nobody was waiting for your startup to arrive. You have to go earn attention. That meant reaching out to people in the e-commerce ecosystem, getting warm introductions where we could, and being very clear about the business outcome we were trying to help with.

A big part of what worked was that we did not sell ourselves as ‘just another AI tool.’ We anchored the conversation around a very concrete problem: brands were spending serious money to drive traffic, but the web experiences on the other side were often generic, slow to launch, and under-optimized. That framing got people interested because it connected directly to revenue.

Once we got a foot in the door, we leaned heavily on hands-on execution. We were willing to do the work, move fast, and prove value quickly. That built trust. Early customers did not just buy software; they bought confidence that we understood both growth and execution.

Two specific things that helped were our advisor network of early supporters who made customer introductions for us and also a free 60-day pilot to prove value before they buy.

Tell us about your primary drivers for growth. What worked for you in the beginning? What’s working now?

In the beginning, our biggest growth driver was founder involvement. We were extremely close to the customer, extremely fast in execution, and very willing to do unscalable things. That let us compress the time between ‘interesting conversation’ and ‘visible value.’ At an early stage, speed matters more than sophistication.

What also worked early was having a clear wedge. We were not trying to be a general AI company. We were solving a painful, expensive, high-frequency problem around landing pages, campaign execution, and conversion. When your pitch is specific, people understand where you fit.

What’s working now is the combination of scalable LinkedIn and cold calling outreach. We have set up drip campaigns and AI GTM workflows to scale these two channels. ROI is clear for our vertical in both these channels.  

What has been your biggest challenge so far, and what did you learn from it? How are you dealing with this issue today?

One of our biggest challenges has been navigating the tension between services and software. Services are incredibly powerful early on. They give you revenue, fast learning cycles, and deep customer intimacy. But if you’re not careful, they can also consume all your energy and quietly stall product development. That balance is harder than it looks.

In the post-AI world, it’s tempting to believe that every service can simply be turned into an AI workflow. The reality is more nuanced. AI has improved dramatically, but it’s still not a silver bullet. Knowing exactly what it can and cannot do reliably is where most teams go wrong.

For us, the real challenge has been figuring out where AI can genuinely replace parts of a service, versus where human judgment still matters. Getting that boundary right and building around it has been one of the most delicate and important problems we’re solving as a company.

Can you tell us about any upcoming developments we can look forward to?

One of the things we’re most excited about right now is the launch of our self-serve product. Until now, a lot of our work has been high-touch, working closely with customers to create and optimize their web experiences. That gave us deep insight into the problem, but the next step is making this accessible at scale.

We’re building a product that allows businesses to generate campaign-specific landing pages quickly and reliably without needing design, engineering, or long iteration cycles. The goal is not just speed, but durability. Pages that don’t just look good, but are grounded in data, context, and what actually converts.

At its core, this will be a fully agentic product. Instead of static tools, you’ll have systems that understand the campaign, pull in the right context, and generate optimized outputs end-to-end. This significantly expands our TAM, because we’re moving from a service-heavy model to a scalable product that any digital business can use. 

What’s the biggest risk you’ve taken as a founder/entrepreneur?

The biggest risk was committing fully to building in a market that is changing in real time. When you start an AI company, the technology moves, customer expectations move, and the narrative around the space moves almost weekly. It would be much easier to wait for more certainty. But that’s also how you miss the window.

For me personally, the risk was choosing the harder path of building something from scratch instead of staying on a more predictable career track. Startups stretch every part of you: your conviction, your patience, your ability to make decisions with incomplete information. But I also think that’s what makes it meaningful.

What was the biggest challenge you had to overcome?

The biggest challenge has been learning how to build conviction without becoming rigid. Early on, as a founder, you are constantly hit with conflicting feedback. One customer wants X, another wants Y. One advisor tells you to scale services, another says kill services immediately. The hard part is deciding what to listen to and what to ignore.

What helped me was getting much more grounded in first principles: where is the pain, who pays for it, what repeats, and what creates unfair value? Once you get clearer on those answers, decision-making becomes less emotional and more strategic.

What tools do you use and recommend?

We’re big believers that the modern stack should be AI-first, especially for software development. Before building anything seriously, we like to prototype quickly using tools like Lovable, even if the backend is rough. Just getting something in front of customers gives you fast, real feedback. Once there’s a signal, we move to actually building using tools like Claude (and Claude Skills), Cursor, and other AI-assisted dev workflows. The exact stack matters less than how effectively you use AI to compress build cycles. We use Fathom to record customer calls and insights from those that are directly incorporated into our product roadmap. 

On the growth side, a lot of the tooling is now agentic as well. We’ve seen strong results using tools like Fal.ai for video generation, ElevenLabs for voice, and bringing it all together in CapCut for ad creation. Most of the stack is now AI-augmented, and the advantage comes from leaning into that early.

What is your favorite quote?

A quote I come back to often is: “The main thing is to keep the main thing the main thing.” It sounds simple, but startups constantly tempt you into distraction. I like reminders that force clarity.

How do you handle work-life balance?

I don’t think early-stage founders always get a neat version of work-life balance. There are phases where the company demands a lot. What I try to optimize for instead is sustainability. That means protecting a few anchors that keep me sane: training, some routine, time with people I care about, and moments where I can zoom out instead of living only in reaction mode.

I’ve also learned that intensity is fine, but chaos is expensive. When I’m able to structure my week even a little better, I show up better both for the company and for my personal life.

What are some trends in your industry that you’re excited about?

I’m excited by the shift from AI as a chat interface to AI embedded inside real workflows. In marketing and e-commerce, the interesting future is not just generating content faster. It’s systems that understand context, use structured inputs well, and help teams make better decisions with less coordination overhead.

I’m also excited by the idea that smaller teams can now operate with much higher leverage. That changes what kind of company can be built, how quickly products can improve, and how much output a focused team can create.

How do you stay motivated during tough times?

I try to stay close to reality. Tough times feel worse when everything is abstract. When I go back to the customer, the product, and the actual next step, things usually become more manageable. Momentum is very underrated psychologically.

The second thing is remembering why I chose this path. I genuinely like building. Even on hard days, there’s something energizing about creating clarity where there was none before.

What’s the best piece of advice you’ve ever received?

One of the best pieces of advice I’ve received is that speed is a strategy. Not reckless speed, but the kind that comes from making decisions, learning quickly, and not overprotecting yourself from imperfect information. In startups, waiting often looks smart but is secretly very expensive.

Can you share a surprising fact about your startup journey that most people don’t know?

A surprising part of the journey is how much of building a startup is really about building your own operating system as a person. People imagine company-building as strategy decks, product decisions, and fundraising conversations. A lot of it is actually emotional regulation, pattern recognition, and learning how not to get pulled apart by noise.

Another thing people underestimate is how much early momentum comes from doing extremely unglamorous things well – follow-ups, customer prep, rewriting positioning, fixing tiny workflow issues, and just staying in the game long enough for compounding to start working.

What is the most rewarding aspect of being a founder/entrepreneur?

For me, the most rewarding part is seeing something go from a vague idea in your head to something real that other people use and value. That transition never stops being satisfying.

What advice would you give to other aspiring entrepreneurs?

Start closer to the pain than to the idea. It’s very easy to fall in love with a concept and much harder to stay obsessed with a real customer problem. But the second route is the one that gives you a business.

Also, don’t wait to look legitimate before you start acting like a founder. Talk to customers early, ship before you’re fully comfortable, and let reality teach you faster than your internal monologue will.