The Five Year Desert to Product Market Fit and a $5.3BN Valuation | Shiv Rao, Founder @ Abridge

Die on the hill for core insights

Rao maintained absolute conviction that healthcare runs on conversations, willing to pivot product features and go-to-market strategy but never the fundamental thesis, even during the five-year wilderness period from 2018 to 2023.

Survive until market timing aligns

Success required simply staying alive until post-pandemic clinician burnout—when 40-50% of doctors reported burnout—coincided with LLM maturity in 2023, creating explosive demand that Abridge was positioned to capture.

Hold strong opinions loosely on expansion

While the core thesis remained fixed, Rao held loosely to opinions about company limits and expansion speed, allowing the organization to scale rapidly and unexpectedly when conditions became favorable.

Don't fight foundation models, ride them

Vertical AI companies must collaborate with and leverage horizontal AI tailwinds rather than compete against them, as attempting to fight foundation models means missing the wave entirely.

Target concentrated upmarket systems early

Healthcare founders should avoid the down-market trap and quickly target integrated delivery networks and academic medical centers where the majority of America's 800,000 practicing clinicians are concentrated.

Own the stack through proprietary training

Abridge generates approximately 40% of model outputs from in-house models trained on daily user edits across diverse care settings, creating defensibility through post-training on proprietary workflow data.

Taste good things to develop good taste

Rao instilled the company value that teams must deliberately expose themselves to diverse influences—from cutting-edge ML research to UI patterns—to create authentic products that feel human-made.

Prioritize founder-investor chemistry

Securing Union Square Ventures' $5 million seed round at a $15 million pre-money valuation came from recognizing shared pattern-matching abilities through music discussions, demonstrating that founder-investor fit matters more than firm brand.

Evolve rapidly through AI epochs

Companies must reorganize significantly to progress through post-transformer, LLM-native, and agent-native eras, adapting both product architecture and internal operations to become the latest variant as fast as possible.