Inside Glimpse: The Hard Pivot That Led to a $35M a16z Round

Most startup pivots happen because the first business clearly failed.

Glimpse was different.

Before raising $35 million from Andreessen Horowitz, CEO Akash Raju and his co-founders already had a working company, real customers, revenue, and backing from Y Combinator.

Then they decided to walk away from it.

What followed was 15 months with no clear product, no obvious market, and no guarantee that another business would emerge.

Eventually, that period led them to an overlooked problem inside some of the world's largest consumer brands: billions of dollars flowing through an extremely manual system of retail deductions, claims, spreadsheets, retailer portals, and financial reconciliation.

In New York, Very Frontier sat down with Raju to understand why the founders abandoned a business that was already working, how they found the current version of Glimpse, and what changed when they went from searching for an idea to raising from one of Silicon Valley's most prominent venture firms.

The conversation covers Glimpse's beginnings at Purdue University, Raju's time at Tesla and Microsoft, the company's original Y Combinator business, the decision to hard pivot, 15 months of uncertainty, and the fundraising process behind its $35 million Series A.

Akash Raju started building companies with co-founders Anuj Mehta and Kushal Negi while they were students at Purdue University.

Before becoming a full-time founder, Raju interned at Tesla and Microsoft.

Their original YC-backed business helped consumer brands place products inside Airbnbs and boutique hotels.

The business had customers and revenue, but the founders believed its growth would eventually plateau.

At the end of 2022, they decided to shut down that direction and effectively start again.

The team then spent roughly 15 months talking to customers and searching for a new problem.

That process led them to retail deductions, a highly manual financial workflow for consumer packaged goods companies.

Glimpse built AI agents that can handle much of the deductions recovery workflow end to end.

The company later signed major enterprise customers and its first seven-figure deal.

In 2026, Glimpse announced a $35 million Series A led by Andreessen Horowitz, with continued participation from 8VC and Y Combinator.

Raju believes raising from a top-tier venture firm ultimately comes down to the market, the founder, and evidence that the company can execute.

Raju's path into entrepreneurship started before Glimpse.

While studying engineering at Purdue University, he completed internships at Microsoft and Tesla.

Tesla proved particularly formative.

Raju joined the factory software team while Tesla was ramping production of the Model 3, placing him close to one of the most difficult periods in the company's manufacturing history.

Even as an intern, he saw software being deployed directly into factory and supply chain operations.

The environment was intense.

There were long hours, overnight work, and constant pressure to solve problems quickly.

But Raju also saw how much leverage a small team could create by building software that was genuinely useful.

It led to another realization.

If he was willing to work that hard for someone else's company, he wanted to see what would happen if he directed that effort toward building his own.

The company that Raju and his co-founders originally built looked almost nothing like Glimpse today.

Their first business helped consumer brands distribute product samples through Airbnbs and boutique hotels.

Imagine checking into a high-end Airbnb and finding coffee from an emerging coffee company in the kitchen, or bedding supplied by a consumer brand trying to get its products in front of potential customers.

The idea gave brands a new physical distribution channel for product discovery.

And it worked.

The founders joined Y Combinator early in their journey, raised outside capital, brought on customers, and generated real revenue.

They were not staring at a failed startup.

That made the next decision substantially harder.

The problem was not whether the business could grow.

It was whether it could grow enough.

Raju describes the distinction as particularly important when building a venture-backed company.

A business can be healthy and still not have the growth profile required to become a venture-scale outcome.

Over time, the founders began to believe their original model would eventually plateau.

There was no single catastrophic event.

Instead, Raju describes it as the type of intuition founders develop when they are operating inside a business every day.

Something in the back of his mind suggested the company would eventually hit a ceiling.

He says he probably felt it for roughly six months before the team finally made the decision.

At the end of 2022, they committed to a hard pivot.

They were effectively starting again.

Walking away from a business is one thing.

Knowing what to build instead is another.

For roughly 15 months, Raju, Mehta, and Negi entered what they jokingly describe as "pivot hell."

There was no clear new company.

No obvious product.

No certainty that their next experiment would be better than the business they had just abandoned.

Their job became talking to customers.

Every day, they tried to understand how consumer brands operated, where employees were spending their time, what workflows were broken, and which problems were painful enough that companies would actually pay to solve them.

Raju describes that period as one of the hardest parts of the company's history.

There was no roadmap to follow.

The founders simply had to continue showing up and learning.

That process eventually became the foundation of Glimpse.

After spending months studying consumer packaged goods companies, or CPG brands, the founders kept encountering a complicated part of selling through retailers.

Deductions.

When a brand sells products through a major retailer or distributor, the amount ultimately paid back to the brand does not always match the original invoice.

Retailers can deduct money for a wide range of reasons, including logistics issues, pricing discrepancies, promotional agreements, missing inventory, damaged products, compliance issues, and other charges.

Some deductions are legitimate.

Others can be disputed.

The problem is figuring out which is which.

Large brands can deal with enormous numbers of transactions across retailers, distributors, documents, invoices, contracts, portals, and internal systems.

Recovering money can therefore involve teams manually downloading documents, cross-referencing information, investigating individual charges, filing claims, and tracking whether disputes were ultimately paid.

It is operationally painful.

But unlike many enterprise software problems, the value of solving it is extremely easy to understand.

Recovering a valid claim puts money directly back into the company's P&L.

This was where the founders saw an opportunity that did not exist in quite the same way when they started their first company.

Traditional enterprise software often gives users better visibility.

It creates dashboards.

It organizes data.

It alerts employees that something needs attention.

But somebody still has to do the work.

Glimpse wanted to go further.

Instead of simply helping employees understand which deductions should be disputed, AI could potentially execute much more of the workflow.

That means retrieving documents, understanding what happened, validating charges, assembling the necessary evidence, filing disputes, tracking claims, handling reconciliation, and pushing information back into a company's systems.

For Raju, that distinction was crucial.

The product was not another dashboard.

It could actually own the workflow.

That was when the founders realized they had found something materially different.

Retail operations provide an interesting environment for AI because so much of the work sits between structured and unstructured systems.

The necessary information might live in an ERP.

Or in a retailer portal.

Or inside a PDF.

Or in an email.

Or in a spreadsheet.

Or across several of them simultaneously.

Historically, humans connected those systems manually.

AI models are increasingly capable of interpreting the documents, understanding the context around a claim, and taking actions based on that information.

That allows software to move beyond helping a person perform the workflow and toward executing more of the workflow itself.

For Glimpse, deductions became the entry point.

The broader opportunity is applying that same model across retail operations.

The contrast between the company's 15 months of searching and what happened afterward is significant.

During the pivot period, the founders were trying to discover what customers truly needed.

Once Glimpse began solving deductions, the signals changed.

The company started landing increasingly large brands.

Growth accelerated.

Eventually, Glimpse signed its first major enterprise contract, including its first seven-figure deal.

Those signals were materially different from simply having people interested in a product.

Customers were willing to deploy the system against financially important workflows and pay meaningful amounts for it.

By the time Raju returned to the fundraising market, the pitch therefore looked different.

He was no longer raising primarily around what Glimpse might become.

He could show what was already happening.

In March 2026, Glimpse announced a $35 million Series A led by Andreessen Horowitz, with continued participation from 8VC and Y Combinator.

The round brought the company's total funding to $52 million.

But the relationship with Andreessen Horowitz did not begin when Glimpse formally started fundraising.

Raju had been building a relationship with the partner involved in the deal for more than a year.

During that period, the investor could watch the company execute.

The founders would make a claim about what they intended to accomplish.

Then they would deliver.

The business would grow.

New customers would arrive.

The numbers would improve.

That sequence gradually reduced the gap between the founders' vision and the evidence available to support it.

Fundraising is often presented as a pitch-deck exercise.

Raju sees it differently.

By the time a company is raising a significant institutional round, particularly from a top-tier venture firm, investors are evaluating multiple layers simultaneously.

Is this the right market?

Can this become a very large company?

Is now the right time?

And is this the team capable of winning it?

The pitch still matters.

But evidence matters more.

In Glimpse's case, the team could point to rapid growth, large customers, enterprise contracts, and direct financial outcomes for those customers.

At the same time, they had to convince investors that CPG and retail operations represented a sufficiently large market for a venture-scale AI company.

Andreessen Horowitz was therefore not simply making a bet on a deductions product.

It was making a bet that AI could become a new operational layer across an enormous industry.

The most interesting part of the Glimpse story may be that its failed first direction was not simply an embarrassing chapter the founders had to overcome.

It became part of the reason the company was investable.

Raju, Mehta, and Negi had already demonstrated that they could build a product, acquire customers, generate revenue, raise capital, and operate a startup.

Then they demonstrated something different.

They were willing to abandon what they had built when they no longer believed it could become the company they wanted to create.

That required accepting a potentially uncomfortable reality: years of effort do not automatically justify spending more years going in the same direction.

The team then spent more than a year operating without the validation that comes from having an obvious product to sell.

By the time they discovered deductions, they had also spent years working with consumer brands.

The pivot had given them something that would have been difficult to manufacture quickly: deep familiarity with an industry that most software founders overlook.

There is an important distinction in Glimpse's story.

Having revenue does not automatically mean a startup has found the right business.

Having customers does not necessarily mean the market is large enough.

And growing does not necessarily mean the company is growing quickly enough to support the outcome its founders and investors are targeting.

Raju's experience shows why product-market fit is not simply a binary question of whether somebody will pay.

The harder question is what happens after they pay.

Does demand compound?

Does the product become increasingly important?

Can customer value be demonstrated clearly?

Can the business expand into larger accounts?

Can the market support a company many times larger than it is today?

For Glimpse, the founders concluded that their first company would eventually fail that test even though it had not technically failed.

From the outside, abandoning a business with revenue, customers, investors, and Y Combinator backing can look irrational.

From inside the business, continuing may have been the riskier choice.

Every year spent scaling a company with a structural ceiling is also a year not spent finding something larger.

The founders could have optimized the original business.

Instead, they reset.

That decision created 15 months of uncertainty.

But it also ultimately led them to a market where customers could measure the value Glimpse created directly in recovered dollars.

That changes the economics of selling enterprise software.

It also changes the fundraising story.

Deductions may be the entry point, but Glimpse is making a broader bet.

CPG companies operate through a complicated network of retailers, distributors, contracts, payment systems, supply chains, documents, and internal software.

Many of the workflows connecting those systems still require significant manual labor.

AI agents create the possibility of turning those workflows into software.

Instead of building applications that wait for humans to push buttons, companies can build systems that interpret information and execute the work.

For Glimpse, recovering revenue is the wedge.

The larger ambition is becoming an operational layer across the consumer and retail industries.

Raju's story is not really about how to pivot faster.

Glimpse took 15 months to find its next direction.

It is about recognizing when persistence becomes attachment.

Founders are constantly told not to give up.

Sometimes that is exactly the right advice.

But persistence toward a mission is not necessarily the same as persistence toward a specific product.

The Glimpse founders kept the team.

They kept their knowledge of consumer brands.

They kept talking to customers.

They kept building.

What they abandoned was the assumption that their first answer had to be the final one.

Several years later, that decision became one of the most important parts of the company's story.

They didn't raise from Andreessen Horowitz despite the pivot.

The pivot may be one of the reasons they became the company Andreessen Horowitz wanted to back.

What does Glimpse do?

Glimpse builds AI software for consumer packaged goods and retail companies. Its platform automates operational workflows including retail deductions, revenue recovery, claims, reconciliation, and cash application.

Who founded Glimpse?

Glimpse was founded by Akash Raju, Anuj Mehta, and Kushal Negi, who met while studying at Purdue University.

What are retail deductions?

Retail deductions occur when retailers or distributors subtract charges from the money they owe a brand. Some deductions are legitimate, while others can be disputed and recovered.

How does Glimpse use AI?

Glimpse uses AI to automate parts of the deductions workflow that traditionally require manual investigation. This can include retrieving documents, reviewing claims, validating charges, collecting supporting information, filing disputes, and reconciling payments.

Was Glimpse originally a different company?

Yes. The founders originally built a business that helped consumer brands distribute product samples through Airbnbs and boutique hotels. The company had revenue and customers, but the founders eventually concluded that its growth potential was limited and decided to hard pivot.

How long did the Glimpse pivot take?

Akash Raju says the team spent roughly 15 months searching for a new direction, talking to customers and studying pain points before arriving at the current version of Glimpse.

How much has Glimpse raised?

Glimpse announced a $35 million Series A led by Andreessen Horowitz in 2026, bringing the company's total funding to $52 million.

Who invested in Glimpse?

The company's investors include Andreessen Horowitz, 8VC, and Y Combinator.

Who is Akash Raju?

Akash Raju is the co-founder and CEO of Glimpse. He studied at Purdue University and completed internships at companies including Tesla and Microsoft before becoming a full-time founder.

Why did Glimpse pivot?

The original company was generating revenue and serving customers, but the founders believed the business would eventually plateau and would not grow at the pace required for the company they wanted to build.