Arish Ali spent years helping people buy things through screens. Then, in 2017, he suggested that the screen's elaborate courtesies might eventually be unnecessary. “The best user interface is no interface,” he said. Coming from someone selling commerce software, it was a pleasingly awkward ambition: build the machinery, then spare the customer the sight of it.
He was interested in voice, artificial intelligence and technology that could recede into the background. Read across his career, that thought offers a useful connecting line. The businesses changed considerably. The question of how much effort software should demand from a person kept returning.
Today, Ali is working on AI applications at Erewhon Labs with his longtime co-founder, Sudha Varadarajan. Before this came Skava, acquired by Infosys, and Swell, a social audio platform that closed in 2025. His career contains both the satisfying acquisition announcement and the less photogenic business of starting over.
A magazine before a machine
His interest in programming began with reading. As a schoolboy in India, Ali encountered lessons in Target magazine before he had used a computer. Code offered an intriguing proposition: write instructions and something would happen. For a child, that is an excellent advertisement. For an engineer, it remains a perfectly respectable motive.
He went on to study electrical engineering at the Indian Institute of Technology Kanpur and electrical and computer engineering at the University of Massachusetts Amherst. Microsoft followed. There he met Varadarajan, who became his wife and business partner. Their subsequent companies would be shared undertakings, even when a headline made room for only one name.
Ali later managed software development at Brience, working on mobile middleware and ways to adapt web content for mobile devices. That experience placed him close to a problem that would occupy him for years: people wanted useful information wherever they happened to be; the software often expected them to sit somewhere else.
The future arrives late
Skava began in San Francisco in 2002. Its name combined the founders' initials, a compact arrangement that saved them the trouble of inventing a mythology. Its early ambitions involved mobile games and a platform for making them. The audience for the idea needed time to catch up.
Ali and Varadarajan adjusted. They used their technology to make games and applications themselves, finding revenue while the broader mobile market developed. Their work included licensed content, with a Spanish for Dummies application among the products Ali later recalled. A company intended to sell a platform had become, for a time, a publisher too.
The shift toward commerce came around 2008. Smartphones were changing what shoppers expected, while retailers were trying to serve them with technology built for desktop browsers. Ali could see a gap between the device in someone's hand and the systems behind the shop. Skava moved into that gap.
He described the approach in deliberately financial terms: choose customers carefully, earn sustainable revenue and work within the budget. Growth could be slower than at a heavily funded competitor. The compensation was control. It is a less theatrical entrepreneurial creed than promising to conquer everything, but it does give a business something to do on Monday.
The shopping cart meets the org chart
As Skava developed, Ali's argument became broader than fitting a shop onto a phone. Retailers wanted the freedom to change their businesses without turning every adjustment into a lengthy software undertaking. Their tools had acquired the unfortunate habit of setting the limits of their imagination.
In 2018, he described customers reducing their requirements to accommodate older commerce systems. A product was available, an idea was ready, but the machinery required negotiation. His interest in modular software came from that frustration. Smaller, separately changeable components could give a retailer more room to experiment.
His writing on microservices laid out the technical obstacle. In a tightly coupled commerce platform, functions share dependencies, so changing one piece can disturb others. Breaking the platform into services offers another way to organize the work. The architecture matters because a retailer still has to keep the shop operating while making improvements.
Ali also drew a distinction between installing a platform and living with it afterwards. In 2017, he put a full Skava implementation at three to six months, depending on scope. His argument was that subsequent changes should take days or weeks, with business users able to update campaigns themselves. A successful installation was supposed to reduce the queue of future requests.
“Retailers need to be obsessed with closing customer experience gaps.”
Arish Ali · 2018
His retail priorities were correspondingly plain: pages that load, products in stock, useful support and delivery that meets expectations. The shopper's attention is occupied by the purchase. A software vendor may admire the complexity underneath; the customer has other plans.
A sale, and a longer life for the software
In April 2015, Infosys announced an agreement to acquire Skava and its affiliate for $120 million, including deferred consideration and a retention bonus. The deal brought a small company's commerce technology into a much larger organization. Financial statements record the acquisition on June 2.
The amount belongs to the transaction, with its various components. A company sale is not a statement of a founder's personal fortune, however tempting the arithmetic may look from a distance. What can be followed more usefully is the work that continued.
Ali remained involved after the acquisition. His remit at Skava included growth, product innovation and client satisfaction, and the company gained access to Infosys's international reach. In September 2021, Infosys introduced Equinox, explicitly describing the commerce platform as an evolution of Skava Commerce.
That gives this chapter an afterlife beyond the sale. A business can change ownership while its technology keeps developing, acquiring new capabilities and a different name. Ali's place in the Equinox story is as co-founder and former CEO of its predecessor. His own experiments had moved elsewhere.
A conversation without a calendar invite
With Swell, Ali and Varadarajan turned to people speaking to one another. The platform launched internationally at SXSW in 2021. Users could post up to five minutes of audio, attach a photograph or link, and receive spoken replies. Participants did not need to be online at the same moment.
That last choice supplied the product's rhythm. Someone could ask a question now and hear an answer later. A conversation could continue around work, travel and everything else people have inconsiderately arranged outside an app. The calendar was allowed to remain its own concern.
The founders were interested in what voice could carry that text and pictures sometimes lost: intention, emotion, the manner of saying something. Varadarajan described their concern about polarization online and the ease with which words could lose context. Swell's proposition put tone back into the exchange while giving speakers time to compose themselves.
- SpeakRecord an audio post
- ListenFind it in your own time
- ReplyAdd your voice later
The India launch followed in June 2021. The team discussed conversations in Urdu, Tamil, Hindi and Telugu, and an ambition to bring more young creators onto the platform. Their proposed revenue model involved subscriptions and premium features. Advertising, they argued, would create the wrong incentives for the kind of exchange they wanted.
The courtesy of taking your time
Swell continued changing. Its 3.0 update in 2022 introduced automatically generated text snippets to help people browse conversations. Audio carried the substance, but a short written preview could help someone decide what to hear. Even a product devoted to voice needed to account for the moment when listening was inconvenient.
Ali's own audio publishing ranged beyond company news. His archive includes a discussion of philosophy of science, recalling a course at IIT Kanpur, and a post on podcasting as audio blogging. There is something revealing in that range: a founder making room for ideas that do not arrive with a product demonstration attached.
In May 2025, he discussed a user's appreciation of Swell's slower pace. He described participants recording replies again to get their tone right. The effort was part of what he valued. A considered response takes time, and a product built around it must accept the consequences.
Swell closed in 2025. The affection of a small, dedicated audience did not keep it running. Ali thanked users publicly and acknowledged how difficult it was to let go of something they had built. The ending belongs beside the ambition; it makes the next beginning less glib.
Back in the workshop
Erewhon Labs brings Ali and Varadarajan together again, with engineers from their earlier ventures. The lab is exploring application-layer AI through prototypes. Its name draws on Samuel Butler's nineteenth-century novel and its questions about people becoming subordinate to technology. An old book has proved useful company for a new set of tools.
Ali's personal description of this phase is modest: “tinkering with AI and trying to understand the future of software startups.” Experience has supplied him with more questions to ask. It has apparently failed to cure the urge to build something.
The visible project list includes a trial for booking local services in South India and work on sharing books and exploring commodity markets. It also marks an idea as canceled. That small word earns its space. An experimental workshop ought to be able to put down a tool without arranging a triumphal procession.
Ali's early programming lessons offered the pleasure of making a machine respond. His later work asks what the machine should require in return. A shopper wants to buy. A speaker wants to be understood. Someone trying new software wants their time treated with respect. Those are demanding customers. He has given himself another chance to meet them.