

Y Combinator
Episodes


Bringing Supersonic Flight To The People

How To Go From Researcher to Founder | Startup School

Engineers Make Great Founders
Former SpaceX engineers Shivani Patel and Chris Markus discuss the transition from high-stakes aerospace engineering to founding AI-driven startups. At SpaceX, the absence of project managers forced engineers to own the entire lifecycle of a project, from problem identification to final implementation. This culture of ...

Infrastructure for the AI Coding Era
Blacksmith is revolutionizing Continuous Integration (CI) by building specialized cloud infrastructure designed specifically for the high-velocity demands of modern software development. Founded by technical "outsiders" who recognized that traditional hyperscalers like AWS are poorly optimized for CI workloads, the com...

Why Everyone Is About To Have Their Own Database
AI agents are fundamentally reshaping software development, shifting the focus from human-centric dashboards to code-first, agent-native infrastructure. Supabase, an open-source Postgres platform, has evolved to support this transition, with agents now accounting for 70% of new activity and over a million databases cre...

AI for the Air-Gapped Economy

AI for City Hall

LLMs That Write Well

Vertical Foundation Models

Wall Street for the Physical Buildout

Own Your Intelligence

Plaid for Government Benefits

Data Center Supply Chain

The $15-an-Hour Expert

Robots Made In America

Startups in a World of Abundant Space Access

The New Full-Stack Designer
AI-driven workflows are fundamentally shifting the design process from manual creation to high-level curation and editing. By leveraging tools like Codex and Claude Code, designers now act as directors who steer AI agents to execute complex technical tasks, such as generating shaders or interactive animations, signific...

Why Inference Matters 12x More Than Training | YC Paper Club
Inference optimization represents a critical bottleneck in scaling AI, with inference flops outweighing training flops by a factor of 12. Achieving high-performance model serving requires balancing time-to-first-token, throughput, and KV cache retention policies. Technical strategies to minimize costs and improve effic...

A Different Approach To Launch Videos
Launching products in the current digital landscape requires moving away from expensive, high-production campaigns toward frequent, authentic "mini-launches" that showcase the development process. By treating demos as mini-stories, founders can build trust and engagement on algorithmic platforms like Instagram and YouT...
Follow this podcast in Podwise
Sign in to get AI summaries, transcripts and mind maps for any episode, including new ones.
