Co-Founder & CEO, Percepta at General Catalyst
Check size: Not applicable. Hirsh leads Percepta, an operating/transformation company, rather than a General Catalyst investment team. Founders seeking GC capital should contact an actual GC investor; enterprises seeking embedded AI transformation can contact Percepta.
Hirsh's operating thesis is that enterprise AI fails when organizations bolt point solutions onto legacy workflows or stop at pilots. Percepta embeds engineers, researchers and product managers inside critical institutions, combining frontier intelligence, proprietary data, domain trust and workforce adoption to redesign operations. The focus is production transformation in healthcare, government, manufacturing and finance—not venture selection or personal check writing.
Do not pitch Hirsh as a General Catalyst investor. Enterprises should approach Percepta with a high-value workflow, the current operational bottleneck, accessible data, accountable users, institutional constraints and a path from embedded discovery to production. Builders interested in employment or research can use Percepta's channels. Venture founders seeking financing should identify a current GC investment partner instead.
Institutions willing to redesign core workflows rather than add an AI demo; high-stakes domains long neglected by modern technology; deeply embedded teams with engineers, researchers and product managers; proprietary institutional knowledge and trusted design partners; production systems that improve work for operators such as nurses or portfolio teams; and ambitious interdisciplinary problems that recent model progress has made tractable.
Point solutions or consulting engagements that remain outside the operating workflow; prototype paralysis; generic AI layered onto legacy processes; technology without domain trust, proprietary data or workforce adoption; and pilots that cannot produce measurable transformation. These are operating-product filters, not GC investment pass criteria.
Start from a consequential workflow and embed technical teams inside the institution until the system changes production operations, not merely benchmark results.
Combine advances in intelligence/models, underlying data infrastructure and workforce adoption so humans and agents can redesign processes together.
Pair frontier technology with proprietary institutional data, domain expertise and trusted customer relationships that an external point solution cannot easily reproduce.
Prioritize healthcare, finance, manufacturing and government problems where better operations create durable public and economic value.
Judge deployment success by whether real operators complete important work better—such as improved care workflows or investment research—not by the novelty of the model.
“Frontier AI is only transformative if it changes how real work gets done.”
— https://www.linkedin.com/in/hirshjain
Enterprise transformation requires frontier intelligence, data infrastructure and workforce adoption, delivered by embedded product, engineering and research teams rather than isolated pilots.
The defining AI deployment challenge is bridging prototype and production in critical industries such as healthcare, manufacturing and government.
+ 63 more investments. View fund →