Grok Bot: What Changes When You Can Build AI Teammates
Lauren Tan and Roman Ugarte discuss Grockbot, a product enabling users to create AI teammates with their own identities and capabilities. They share how AI agents are transforming software engineering and business operations, demonstrating workflows where engineers ship thousands of PRs monthly by leveraging specialized bots integrated into team communication tools.
Summary
Grockbot is a product that allows users to create AI teammates—autonomous agents with distinct identities, tools, and long-running goals. Lauren Tan and Roman Ugarte, who incubated Grockbot over seven weeks, explain that the product emerged from observing how AI coding transformed engineering workflows. They emphasize three core principles: bots should be organized as key operative units, have access to the same tools humans use, and run continuously to be proactive.
The product launched with surprising internal reception across both AI-forward and AI-skeptical users. Real-world deployments show bots automating functions across departments—marketing teams give bots access to Google Ads and Facebook Ads with budgets, support teams use bots to triage and resolve tickets, and sales teams have created shared specialist bots for lead research and deal management.
Lauren discusses her personal workflow where she uses specialist bots (like her "Potato" bot running in "Lauren mode") integrated into Slack rather than siloed in a separate app. She collaborates with colleagues on features entirely within Slack threads, eliminating the need to manually copy context between tools. Her engineering bots merge their own pull requests, and she ships over 4,000 PRs monthly by establishing strong code guardrails, static analysis, and lint rules that encode quality standards into the codebase rather than relying on expensive token-intensive code review.
Roman highlights how different organizational structures are emerging: some users prefer one comprehensive bot, others maintain specialist bots organized horizontally (one per customer) or vertically (one per function). Small businesses particularly embrace Grockbot out of necessity—they can now automate accounting and administrative tasks that would be too expensive to hire for.
The team addresses challenges of AI agents working in multiplayer environments. Bots struggle with social awareness and context about who holds decision-making authority in team discussions. They're exploring how bot identities, permissions, and authentication should work—drawing parallels to how companies onboard human employees with role-based access that expands as trust builds.
A critical insight emerges about future software engineering: as execution becomes abundant and cheap, the bottleneck shifts from engineering bandwidth to intent specification and product vision. Companies will need to excel at defining objectives and KPIs for agents rather than managing execution details. Lauren's approach to review emphasizes an "inverted pyramid"—heavy investment in codebase architecture, static analysis, and lint rules reduces the need for expensive agent review cycles. She's also developed Pstack, a set of opinionated skills teaching agents her preferred engineering practices, which she's open-sourced for others to adopt and adapt.
The conversation touches on the emerging concept of AI-native companies where traditional middle management roles dissolve and individual contributors gain autonomy to build features end-to-end. New roles emerge: domain experts become bot builders and managers, and a "player coach" role teaches the organization AI best practices. Human judgment and taste remain critical—despite velocity, the team maintains high standards by requiring internal product-market fit before shipping features, resisting the temptation to deploy every possible feature.
Key Insights
- Lauren ships over 4,000 PRs monthly by encoding quality standards into the codebase through static analysis and lint rules rather than relying on expensive token-intensive code review, demonstrating that investment in architecture and tooling is more efficient than iterative agent review cycles.
- Grockbot received surprisingly uniform internal enthusiasm from both AI-forward and skeptical employees because it mapped naturally to how people already work, suggesting that successful AI product design fits existing workflows rather than requiring behavioral change.
- Small businesses are the most enthusiastic early adopters of Grockbot because they face hiring constraints—they can now use bots to automate bookkeeping and administrative functions that would be too expensive to hire for, suggesting AI addresses different pain points across company sizes.
- As execution becomes abundant through AI, the organizational bottleneck shifts from engineering bandwidth to intent specification—companies must excel at defining objectives and KPIs for agents rather than managing execution details, fundamentally changing what leadership and product work entail.
- Lauren's workflow demonstrates that by integrating bots directly into Slack where work happens and giving them access to external tools, engineers can collaborate with AI agents as transparent teammates in public channels, enabling other team members to observe and replicate effective practices.
Topics
Transcript
[0:00] Grockbot is meant to be this product where you can create AI teammates. >> They have their own identity, their own telephone number, credit card, like basically almost anything uh a human can do. I think we would want bots to do. I sort of see myself as the explorer, you know, like trying to figure out what what does software engineering look like 6 months to a year from now. Welcome back to another episode of the Light Cone. Today we're joined by Lauren [0:32] Tan and Roman Ugarte, the team behind Grockbot. Roman and Lauren incubated the product and are on the core team that took it from the first line of code to launch in just…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Y Combinator
We’ll Be Inferior to Our Grandchildren—and That’s Great
A discussion about scaling societies and the relative ratio of good versus bad behavior, with one participant expressing concern that future generations will be intellectually superior to current humans, similar to how humans are vastly superior to ants.
YC Debates: Should We Slow Down AI?
Jacob Coxin argues governments should substantially intervene to slow frontier AI development, citing rapid capability growth (superhuman AI researchers by 2027), alignment risks, and misaligned corporate incentives. Robin Hanson counters that typical technological adoption patterns suggest slower real-world impact, that regulation is dangerous without concrete problems, and that liability-based approaches are preferable to centralized global regulation.
How To Build The Future: Eric Glyman, Co-CEO of Ramp
Eric Glyman, co-CEO of Ramp, discusses how the company evolved from Parabus to build a $60 billion financial operations platform, and how Ramp is leveraging AI agents and internal tools like Inspect, Research, and Glass to transform organizational productivity and reshape the future of work.
Grok Bot: What Changes When You Can Build AI Teammates
Lauren Tan and Roman Ugarte discuss Grokbot, a platform for creating AI teammate agents with their own identities and tool access that can work autonomously. The conversation covers how AI agents are transforming software engineering and non-technical work, with Lauren showcasing her 4,000+ PRs/month workflow using specialized bots in Slack and describing the shift from execution-constrained to execution-abundant organizations.
Is AI Now Good Enough To Cure Cancer? | YC Paper Club
The YC AI for Bio Paper Club event explores how artificial intelligence is transforming cancer research and biological discovery through foundation models, personalized genomic analysis, and autonomous research systems. Speakers demonstrate that cancer is becoming a data problem solvable through AI agents, DNA language models like EVO2 are enabling efficient protein and genome design through few-shot prompting, and validated lab results show these approaches are moving from theory into clinical practice.