How a 26-year-old raised $108M by solving an urgent problem

Executive overview

Most early-stage startups chase the idea they believe in rather than the problem the market is screaming for. Adit Abraham and his co-founder at Reducto learned this the hard way: they abandoned a long-term memory API product despite genuine interest because urgency was missing — customers would pay $50–100/month but never prioritised it.

A throwaway feature for file parsing drew immediate inbound demand. They followed it.

The fastest path to growth is solving the problem customers are actively searching for a solution to right now — not the one they'll eventually need.

Pivoting to the right problem

• Long-term memory API (Rememberall) attracted interest but low urgency — a "nice to have"
• File management added as a minor feature generated immediate excitement and inbound purchase requests
• Customers asked "Is this a hosted API? Do you have a Stripe link?" — pulled rather than pushed
• Switched focus entirely to document parsing; Reducto now processes over a billion pages

Early growth tactics

• Published a public playground so any company could test Reducto on their hardest documents
• A trillion-dollar enterprise booked a demo after uploading documents that had failed every other vendor
• Being public about the product's capabilities replaced the need for marketing claims
• Companies that initially refused to trust an early-stage team came back inbound in year two

Customer intimacy as a growth strategy

• Individual Slack channels with every customer from day one
• Direct phone numbers shared; co-founders on call for urgent production issues
• Treated trust from early adopters as a debt to repay through personal availability
• Design-partner relationships: customers surface the hardest edge cases, Reducto iterates daily
• Production financial and healthcare data unavailable on the internet — customer access is a competitive moat

Doing the unsexy work

• Adit personally labelled document boxes when the data-labelling team lacked accuracy
• Manually set up every Stripe subscription before billing was automated
• When landing the first major on-prem enterprise deal, the team worked until exhausted, slept briefly, and repeated
• No role boundaries in early team: engineers did customer support, ops staff labelled ML data

Choosing investors over firm brands

• Partner relationships matter more than firm reputation — they're decade-long partnerships
• Evaluate investors by their behaviour when things go wrong, not when things are going well
• Seed investor stepped out of a Broadway show during an OpenAI outage to connect Reducto with the CPO
• Early-stage founders should prioritise individual partner fit over tier-one brand recognition