Jay Hack

Jay Hack joins Keith to discuss the future of knowledge work as agents become the primary way work gets done. The conversation covers Jay's path from Palantir and Mira to Codegen and ClickUp, why context is the practical ceiling on agent performance, ClickUp's super agents and VibeUp, first-party data integration, generative UI, multi-agent work architectures, and brain-machine interfaces as a potential productivity step-change.

Jay Hack, Head of AI at ClickUp and co-founder of Codegen, talks about how AI is reshaping productivity, software, and the future of work. The conversation traces Jay's path from Stanford AI work to startups and ClickUp, then covers myths about AI, hallucinations, AI-native companies, agents, ClickUp's AI strategy, tools beyond ChatGPT, and why future work will involve humans collaborating with AI systems.

Jay Hack interviews Chroma founder and CEO Jeff Huber about why context engineering remains central to AI engineering and how search infrastructure is changing for AI-native applications. They cover context rot in long-context models, regex versus semantic search, LLMs as query planners, Chroma's indexing of open-source package ecosystems, memory systems, auditability, agent learning, and the lack of innovation in AI user experience.

Jay Hack interviews LangChain co-founder and CEO Harrison Chase about LangChain's evolution from a Twitter prototype into an agent infrastructure company. They discuss deep agents, planning, subagents, file systems for long-running tasks, Claude Code, UX and workflow differentiation, sandboxing, agent runtime primitives, memory, and where startups can still compete as model providers push into coding.

Jay Hack interviews bloop co-founder Louis Knight-Webb about Vibe Kanban and the orchestration layer for running multiple coding agents in parallel. The conversation covers the move from code search and COBOL modernization to agent orchestration, Git worktrees, sandboxing, using Vibe Kanban to build itself, human review and planning bottlenecks, GitHub and project-management integrations, enterprise versus startup workflows, and which tasks are a good fit for coding agents.

Jay Hack interviews Graphite co-founder and CEO Merrill Lutsky about how AI code generation is stressing traditional development workflows and making code review the bottleneck. The discussion focuses on engineering workflow changes, pull request review, review quality, and the operational implications of AI-generated code for teams shipping software.

Jay Hack discusses Codegen as an AI-assisted developer platform for refactoring large enterprise codebases at scale. The interview covers the code generation landscape, AI hype versus practical value, why refactors are a strong fit for AI, when to improve an existing codebase instead of rewriting it, Codegen's graph-based SDK and static analysis approach, constraining AI to reduce hallucinations, customer time savings, and platform engineering teams as the initial ICP.

Jay Hack presents Codegen during the Founders You Should Know February 2024 showcase. The segment introduces Codegen as an AI engineer focused first on code generation for problem-ticket resolution, alongside the company's hiring needs for applied algorithms and staff engineering roles.

Jay Hack, founder and CEO of MIRA Beauty, on intelligent sampling — how brands can finally close the first-party data gap with consumers by combining smart sampling experiences with computer vision so that every interaction teaches the brand something about who the customer is and what they want.

Jay Hack speaks at "The Future of eCommerce for Beauty and Skincare" about how AI, computer vision, and personalization are reshaping product discoverability online. The session covers why product discovery is the central problem for beauty e-commerce, what better recommendations look like in practice, and how MIRA's approach builds a feedback loop between consumers, content, and brands.