Akshat Tyagi has a rule for customer calls: someone who knows more than the founder should correct the founder, right there, while the customer is listening. There is no ceremonial interval in which the chief executive’s explanation must be allowed to survive. A product manager with a better grasp of the work should interrupt. The client has a project to deliver. Protecting the founder’s dignity is an unrequested extra.
By August 2026, Tyagi was writing about this as a practical requirement of enterprise partnerships. His company, Metaforms, had confused customers by explaining its internal product divisions. Data validation and the coding of open-ended answers occupied separate boxes inside the business. A customer saw one process and wanted one usable result. The architecture lesson was wasting everybody’s afternoon.
It is an unusually revealing place to begin with a founder who could introduce himself through a funding announcement. Tyagi is the CEO of an AI company that raised a $9 million Series A in July 2025. Yet the more interesting personal thread runs through his willingness to reconsider the lesson, the teacher and, occasionally, his own answer. He has been occupied with education for rather longer than he has been selling research software.
A good student questions the exam
Long before Metaforms, Tyagi was writing against the habits of formal schooling. His 2017 essay on the attention given to examination toppers objected to treating education as a public ranking exercise. He was the author of Naked Emperor of Education, already prepared to take an argumentative title into the world. He questioned what the result measured and what the surrounding applause encouraged.
His later account of starting edhack made the objection personal. He described doing well academically while pursuing much of his learning independently from age fourteen. The system had rewarded him. That made his criticism more complicated than a complaint about losing: he wanted others to have access to the freedom and resources he had enjoyed. Reading widely, writing and visiting alternative learning spaces were part of his own education.
There is a pleasing accident in the next chapter. Searching online for himself, he found Akshay Tyagi, whose name differed by one letter. Akshay had written about leaving engineering college. Akshat emailed him. The correspondence became a working relationship, and in May 2018 they created Unboard.org, gathering resources for people considering different routes through learning and work. A search for his own name had produced a collaborator. Vanity, on this occasion, proved administratively useful.
Edhack followed as an attempt to make independent learning more accessible. Tyagi also became a Rajeev Circle Fellow in 2019. His publisher’s biography records an earlier content-company sale in 2014, placing entrepreneurship well before the present AI business. The sequence matters: his working life already included building something, handing it on, and beginning again.
- Question the rankingWrites about education and examination culture.
- Cross the subjectsCo-authors Now That We’re Here.
- Learn the customer’s jobMetaforms studies agency operations.
- Make room for correctionWrites about enterprise partnerships.
A book with room for more than one subject
In April 2021, Penguin published Now That We’re Here: The Future of Everything, co-written with Akshay. Its scope moved across technology, democracy, design, economics and data. The premise was that preparing for a changing world required looking across subjects and reconsidering how people learn. It was a large brief for a book, and a recognisable continuation of his earlier interests.
The connection to his later company is a question running through both projects: how do people acquire enough understanding to make a useful decision? In the book, that question crosses disciplines. In market research, it meets a questionnaire, a deadline and a bill. The scale of the ambition can be broad; the next piece of work still needs to be specific. That tension gives Tyagi’s career more continuity than the jump from education to enterprise AI might first suggest.
Three pivots, then an apprenticeship
Metaforms, previously called WorkHack, became another education. Tyagi and co-founder Arjun S did not start as veterans of research operations. Their early search involved changing direction, selling across different industries and building things buyers did not want. A later fireside conversation with investor Manav Garg was advertised around three pivots. Tyagi’s own affectionate description of that stretch was “pivot hell”. The phrase leaves little room for a tidy founding myth.
The company’s account dates its real introduction to market research to late 2024. The team paid a former agency operator to walk through his inbox and spreadsheets. It is a wonderful curriculum for someone trying to understand a business. The inbox contains the interruptions. The spreadsheet contains the obligations. Between them sits much of the work a polished sales presentation leaves out.
Tyagi had spent years thinking about learning outside conventional classrooms. Here was a concrete instance: paying an experienced practitioner to explain the job before attempting to automate it. The expert’s working materials became the lesson. The founders had to sit on the receiving end of an explanation, which is a useful position for anyone about to offer one.
Arjun’s engineering account adds a second piece of that apprenticeship. The team established a Mumbai office dedicated to domain expertise, bringing in experienced survey programmers and data-processing specialists. Those practitioners helped construct examples and test the agents against demanding cases. Their knowledge included habits and exceptions that were poorly represented in public online material. Hiring expertise was part of learning how to build the software.
The work behind the question
A market-research survey has an ungainly life behind the neat screen a respondent sees. Someone translates questions into code, checks routing, coordinates sample providers and makes sure the data meets the project’s requirements. Metaforms builds agents for those jobs. Tyagi’s stated ambition is to reduce the operational load so agencies can accept more work and make professional research accessible to more businesses.
The daily constraint gives his learning story some teeth. An agency cannot accept a system merely because a demonstration is entertaining. Its staff must review the output and remain responsible for the result. In his writing about selling enterprise agents, Tyagi puts validation and reliability ahead of the headline time saving. A bad output on a client project can damage the relationship on which the business depends. Speed has to earn permission to matter.
His thinking about interfaces follows the same concern. He has argued that experienced users should be able to edit the same work an agent creates. A specialist who knows the precise change should have a direct route to making it, rather than composing another request and waiting. The principle respects an inconvenient truth for software makers: the person using the product may understand the job better than the product does.
“I need to check”
Akshat Tyagi’s suggested answer before making a confident guess
What changed when customers arrived
The commercial evidence arrived in 2025. In July, Metaforms announced the $9 million round led by Peak XV Partners, with Nexus Venture Partners and Together Fund participating. At that point, six months after commercial launch, the company reported four of the world’s twenty largest research agencies as customers and more than 1,000 surveys processed per month. Strat7 was among the named agencies.
These figures locate a change in Tyagi’s responsibilities. Getting a first product used is one task; maintaining a service on which agencies rely is another. Together Fund had backed Tyagi and Arjun from the seed stage, giving the relationship with Manav Garg a history that preceded the larger round. The people who listened during the uncertain period were still there when the announcement became easier to explain.
In July 2025, Tyagi also spoke about opening a New York office and hiring there. His LinkedIn profile now places him in New York. His 2025 retrospective described the city as the company’s primary home, reported a team that had grown from eight to thirty across APAC and North America, and counted more than 10,000 surveys going live through Metaforms during the year. Those are his year-end figures, a snapshot of the business at that stage.
The retrospective also reported reaching multi-million-dollar annual recurring revenue from zero over twelve months, without supplying an exact figure. For someone whose earlier company story included products customers did not want, the contrast is substantial. There were now buyers, repeated use and a team with work to divide. Finding a market had created a fresh set of things to learn.
31 July 2025
reported as customers then
People who will argue with him
Tyagi’s account of two early colleagues makes that transition less abstract. Santoshi Meesala and Jatin Pandey joined in founder’s-office roles, a title broad enough to cover whatever might be necessary before customers and revenue arrived. Over time, their responsibilities became more defined: sales development for Santoshi, customer success and community for Jatin. He described their development with obvious affection.
The telling detail was his appreciation that they challenged him when they thought he was falling short. That sits comfortably beside the instruction to correct the founder on a customer call. Both make disagreement part of doing the work. An early employee needs the freedom to say that the company’s explanation, promise or behaviour needs fixing, even when the explanation belongs to the person in charge.
Read together, his public accounts suggest a founder who values correction enough to keep returning to it. That is an interpretation of the work he describes, rather than a claim that every day at Metaforms follows the principle perfectly. He writes openly about mistakes precisely because they have happened. An organisation can announce that it listens in a sentence. Its customers discover the truth over several conversations.
There is a satisfying turn in this education. The young critic of ranking systems became an author, an education entrepreneur and then a founder whose customers grade the output themselves. Their tests are specific. Does the survey work? Can the specialist make the necessary edit? Did the team understand the correction? Tyagi’s next lesson arrives with the next project. He has built himself a classroom where the examiners have deadlines.