The Great Loops Debate — Dex Horthy, Geoff Huntley, Ian Livingstone, Greg Pstrucha, moderated by Allie Howe
◆Allie Howe · Dex Horthy · Geoffrey Huntley · Ian Livingstone · Greg Pstrucha
AI Engineer World's Fair 2026 · 1:00:16
Application monitoring and AI debugging
Sentry builds application monitoring software that helps developers diagnose failures and improve code performance. Its platform connects errors, logs, session replays, traces, profiles and metrics, letting teams investigate slow requests and failed interactions with the surrounding application context. Seer, its AI debugging agent and code reviewer, uses production telemetry and code history to explain failures, generate patches and identify potential regressions before changes ship. Integrations with GitHub, Slack, Jira and Linear bring that context into developers’ existing workflows.
Founded by David Cramer and Chris Jennings, Sentry grew from an open-source project for handling application errors. Milin Desai joined as CEO in 2020 and continues to lead the company. Its engineering includes Snuba, a service built on ClickHouse that ingests data from Kafka and provides a query optimizer and data model for searching and aggregating telemetry. Originally replacing a combination of Postgres and Redis for error analysis, Snuba now supports much of Sentry’s time-series functionality.
In 2026, Sentry reported use by 200,000 organizations across 146 countries. Its 2022 $90 million Series E brought total funding to $217 million at a valuation above $3 billion. Sentry acquired code coverage provider Codecov in 2022; Harness acquired Codecov from Sentry in June 2026, moving that business to a separate owner.
◆Allie Howe · Dex Horthy · Geoffrey Huntley · Ian Livingstone · Greg Pstrucha
AI Engineer World's Fair 2026 · 1:00:16
AI Engineer Europe 2026 · 17:05
Affiliations reflect their AIE appearances, not necessarily current employment.
Start here for a daily engineering workflow informed by tracked prompt usage, showing where comprehension tasks fit alongside code generation.
Priscila Andre de OliveiraAI Engineer Europe 2026
Learn how monitoring and error context can reach Cursor and VS Code through MCP, and why Cramer treats the protocol as a pluggable agent architecture.
David CramerAI Engineer World's Fair 2025
Use the multithreaded counter example to examine the limits of permissive typing and testing when agents produce code that runs but may be unsafe.
Daniel SzokeAI Engineer Europe 2026
Moderator Allie Howe leads an Oxford-style debate between Ian Livingstone and Geoffrey Huntley, who argue that disciplined coding-agent loops already deliver practical autonomy, and Dex Horthy and Greg Pstrucha, who challenge the gap between loop hype and reliable engineering outcomes.
Allie Howe · Dex Horthy · Geoffrey Huntley · Ian Livingstone · Greg PstruchaAI Engineer World's Fair 2026
Priscila Andre de Oliveira demonstrates a Catch Me Up repository-comprehension skill using Claude Opus, recommending that developers understand an agent’s findings before moving into planning or implementation.
David Cramer uses Sentry’s MCP server to examine remote transports and OAuth for B2B SaaS, including how excessive API calls and unconstrained responses can push costs onto users.
Daniel Szoke presents Rust’s compile-time rejection of data races as a mechanism for repairing unsafe generated code. The joint debate featuring Greg Pstrucha broadens the discussion to verification, code review, CI/CD, and coordination between agents.
Affiliations reflect each recorded session, not necessarily current employment. The Great Loops Debate is a joint discussion featuring Greg Pstrucha alongside panelists from outside Sentry.