Verloop Spent a Decade on the Chat Window. Nurix Just Bought It

The best startup ideas rarely arrive as ideas. They show up as chores. In 2015, two founders in Bengaluru, Gaurav Singh and Piyush Mishra, were running a concierge app called Magictiger and quietly drowning in it. To keep customers happy, they had built a chat operations team of more than 100 people. That team spent its days answering the same handful of questions - order status, refunds, hours, availability - over and over. So the founders started automating the boring parts for themselves. Then they looked up and realized every other consumer business had the exact same problem.

That side effect became Verloop.io. Eleven years later, in July 2026, it was acquired by Nurix AI, the enterprise AI company founded by Myntra and Cult.fit veteran Mukesh Bansal. The deal folded Verloop's chat automation into Nurix's voice-agent platform. It is a small, tidy transaction on paper. It is also a near-perfect snapshot of where the whole conversational AI industry is heading.

01 — The OriginThe chore that became a company

Singh was not a first-time founder. Before Verloop he had co-founded ventures including GoDeliver and Stalk Ninja, and worked at TCS and Vizury. Mishra was a developer who had built alongside him. Neither had set out to build a customer-support company. They backed into it because they were living the pain.

The insight was almost embarrassingly simple: if your best support agents spend most of their time on repetitive questions, the software should handle those, and the humans should handle the rest. Verloop's early product was a chatbot builder that could answer routine queries, qualify leads, and - the part that actually mattered - hand off cleanly to a person when it hit the edge of what it knew.

Early customers were Indian consumer names - Nykaa, Apollo Munich, Portea, Livpure. In 2017 the company went through Y Combinator's Track'17. By 2018 it had raised roughly $3 million led by IDFC Parampara Fund and was doing about $960,000 in revenue. Modest numbers, but the shape of the business was already there.

02 — The GrindThe unglamorous years

What happened next is the part most founders skip when they tell their story, because it does not photograph well. From roughly 2018 to 2022, Verloop just kept building. A Series A of about $5 million led by Alpha Wave Global in 2020. Expansion into the UK, UAE and Australia. More languages. More integrations. By the time the numbers were public again, the platform supported 80-plus languages and was handling more than 20 million customer conversations every month across banking, fintech, retail and telecom.

That $15 million total raise is worth sitting with. Verloop reached an acquisition on a fraction of the capital its American counterparts burned through. It is a quiet argument for capital efficiency in a category that has otherwise become a firehose of venture money.

The moat was never the model. Anyone can call an LLM now. Verloop's real asset was the messy stuff underneath: a decade of real customer conversations in dozens of languages, deep enterprise integrations, and the trust of banks and telecoms that do not hand their customer data to just anyone. That is the kind of thing you cannot fundraise your way past. You have to sit in the mess and collect it.

03 — The LandscapeSame category, four different bets

To understand why Verloop got bought instead of buried, it helps to look at who else is in the room. Conversational AI for customer support has splintered into distinct strategies, and each company is a bet on a different answer to the same question: what do you own when the model becomes a commodity?

Decagon is the maximalist US bet: raise huge, sell enterprise AI agents, and let valuation do the talking. In early 2026 it raised $250 million and tripled its valuation to $4.5 billion in under six months. Rasa is the opposite instinct - an open, self-hosted framework for teams that want control and no vendor lock-in. Botpress chose the cloud-builder path, easy to start, hosted by default. Retell AI narrowed all the way down to low-latency voice phone agents billed by the minute. Forethought, Quiq and Teneo.ai each carved their own enterprise niches.

Verloop's bet was distribution and language. Not the biggest valuation, not the purest open-source story - just deep roots in the markets, India and the Middle East, where support automation is most needed and least served. That focus is exactly what made it worth acquiring.

04 — The DealVoice buys chat

Nurix AI is barely two years old. Founded in 2024 by Mukesh Bansal and Abhishek Asawa, it builds voice and chat agents for enterprise customer service, sales and operations, with clients including super.money, Myntra, Cult.fit and Aditya Birla Capital. Its voice platform is called NuPlay. What it did not have, at Verloop's depth, was chat.

So it bought it. The undisclosed July 2026 deal gives Nurix a combined platform that can automate a customer interaction whether it starts as a phone call or a WhatsApp message. Gaurav Singh did not cash out and disappear; he joined Nurix's leadership team to run product strategy, enterprise go-to-market and AI-agent development. The combined operation is set to automate more than 20 million monthly interactions.

Read at the industry level, this is the story of the year. The standalone-tool era of conversational AI is ending. Voice companies are buying chat companies. Chat companies are adding workflow. The winners are assembling full-stack agents that own the whole conversation, and consolidation is how they are getting there fast. Verloop-into-Nurix is one deal in a wave of them.

05 — The TimelineEleven years, briefly

06 — The TakeawayWhat you can steal from this

Strip away the acquisition headline and Verloop leaves behind a few reusable lessons. Watch what your best people do all day - the repetitive, expensive work is usually the product hiding in plain sight. Build in a boring, underserved market and let compound interest do the rest; India and the Middle East were not where the biggest valuations lived, but they were where the need was real. And keep asking the only question that survives the AI hype cycle: when the model is free, what do you still own?

Verloop answered that question for eleven years, one chat window at a time. That is why it had something worth buying.

FAQQuick answers