applied-ai
(11)
Pypestream Bets the Enterprise on AI That Actually Finishes the Job
The New York AI company crossed 50 million monthly interactions for Fortune 500 clients in 2026, staking its pitch on agents that resolve problems rather than just talk about them - and drawing a line between itself and rivals like Teneo.ai and Parloa.
Aakash Mahalingam
Aakash Mahalingam is the cofounder and CEO of Canary (YC W26), a startup building an AI QA engineer that reads your source code to catch broken user flows before they hit production. Before Canary he built Windsurf's enterprise product and helped push its self-serve plan to roughly $20M ARR in six months. An ICPC APAC finalist with earlier stints at AWS and Morgan Stanley, he now works out of Mountain View on the problem of testing keeping pace with AI-accelerated coding.
Blueshoe
Blueshoe is an AI-native law firm that pairs licensed attorneys with proprietary legal reasoning software to make quality legal help affordable and always-on for individuals and small businesses. Founded in 2025 by Harvard Law graduate Casey O'Grady and MIT-trained engineer Kai Yee Wan, the Y Combinator-backed company started as a legal-reasoning research platform piloted at Harvard, Yale, and Columbia law schools, then extended into a full-stack firm covering employment, housing, consumer, personal injury, and product-liability matters without billable hours.
Percepta
Percepta is a New York-based AI transformation company, wholly owned and incubated by venture firm General Catalyst, that embeds teams of researchers, engineers, and product managers directly inside large enterprises to move them past AI pilots and into AI-native operations. Founded in 2025 by CEO Hirsh Jain (a former Palantir SVP) and General Catalyst's Hemant Taneja, the company pairs applied engineering with a frontier research lab to rebuild core workflows in healthcare, financial services, manufacturing, and government - working closely with founding partners Anthropic and AWS.
Fleet AI, Inc.
Fleet AI is an applied research lab building training gyms for AI agents - high-fidelity simulation environments that model real work so frontier models can practice, fail, and learn under human supervision. The company partners with AI labs, hyperscalers, and enterprises to develop the next generation of embodied, drop-in virtual assistants.
Hirsh Jain
Hirsh Jain is the co-founder and CEO of Percepta, a General Catalyst Transformation Company that embeds AI engineers, researchers, and product managers directly inside large enterprises to move them past pilots and into AI-native operations. He spent seven years at Palantir Technologies as a senior vice president leading its healthcare and civilian-government business, and previously worked at Google, Jane Street, and the U.S. Department of Defense. A Harvard computer-science and mathematics graduate and published machine-learning author, he launched Percepta in 2025 with General Catalyst CEO Hemant Taneja and a founding team drawn from Palantir, MIT, Meta FAIR, Google, and Citadel. Shortly after launch, Palantir sued him and other former employees, putting the New York startup at the center of one of the year's most-watched AI talent fights.
QBurst
QBurst is a global digital product engineering and consulting firm founded in 2004 in Trivandrum, India. It builds custom software, cloud platforms, data and AI solutions, and digital experiences for a blue-chip client base across the US, Japan, Europe, the Middle East, and South Africa. Now majority-owned by Multiples Alternate Asset Management after a ~USD 200 million 2025 deal, QBurst employs roughly 3,500 people across more than 20 cities and positions itself as a 'High AI-Q' partner blending human expertise with intelligent technology.
David Cheng
David Cheng is a General Partner at Costanoa Ventures, an early-stage venture capital firm in Palo Alto focused on data, AI infrastructure, cybersecurity, and fintech. With over a decade of investing experience - including nine years at DCM Ventures, where he became the youngest VP in the firm's 30-year history - Cheng has backed breakout companies including EvenUp (now valued at $2B+) and Brigit (acquired for up to $460M). A Forbes 30 Under 30 honoree and Wharton graduate, he is also co-founder of Coreline Ventures and publishes the 'Earned Intuition' newsletter on Substack, sharing lessons from a decade of early-stage investing.
John Cowgill
John Cowgill is a General Partner at Costanoa Ventures, a Palo Alto-based early-stage venture firm with over $2 billion in assets under management. Promoted to GP in September 2024, he has spent nearly a decade building the firm's cybersecurity and space practices, backing category-defining companies like AppOmni, Cyberhaven, Muon Space, and Kepler Communications. A Northwestern University alumnus and former McKinsey consultant who describes himself as a 'wayward Kentuckian and theater school dropout,' Cowgill invests in Seed and Series A startups across applied AI, AI infrastructure, cybersecurity, space, and B2B fintech—leading with a founder-first philosophy anchored in conviction, humility, and grit.
Josh Coyne
Josh Coyne is a Partner at Kleiner Perkins, one of Silicon Valley's most storied venture capital firms, where he has backed category-defining companies including Figma, Rippling, Loom, UiPath, and Synthesia since joining in 2017. A summa cum laude Boston College graduate in Computer Science and Finance, Coyne cut his teeth at Qatalyst Partners advising on multi-billion dollar tech M&A before pivoting to backing founders himself. He focuses on enterprise software, applied AI, fintech, and data infrastructure - and co-hosts Kleiner Perkins' Grit podcast, where he probes founders on product-market fit and go-to-market strategy. Originally from Cleveland, fluent in French, and apparently devoted to Chick-fil-A.
Eugene Yan
Eugene Yan is a Principal Applied Scientist turned Member of Technical Staff at Anthropic, where he bridges cutting-edge AI research with production-scale systems. Formerly at Amazon for five years building real-time recommendation and LLM-powered systems for Kindle and Search, Eugene is equally well-known for his prolific writing: 209 blog posts, 420,000+ words published, and a newsletter with over 11,800 subscribers. His open-source repository applied-ml on GitHub has become a canonical reference for teams shipping machine learning in production. He lives in Seattle, snowboards on weekends, and writes like someone who actually wants you to understand.