# Ben Hylak

Markdown copy of https://realalpha.app/people/ben-hylak/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

Real Alpha has cut 4 moments of Ben Hylak (Co-founder & CTO, Raindrop) talking, from an interview on MTS published Nov 8, 2026.

- Role: Co-founder & CTO, Raindrop
- About: Built Raindrop’s tools to detect, investigate, and fix failures in production AI agents.
- On X: https://x.com/benhylak

## What Ben Hylak said

- [Simulations replay production traces to test agent changes](https://realalpha.app/s/simulations-replay-production-traces-to-test-agent-changes/)
  MTS · Nov 8, 2026 · 3 min
  > “most teams right now are not able to iterate on their agent in a meaningful way without shipping to production.” — Ben Hylak · [10:21](https://realalpha.app/v/s_wkGlV-wr4/?t=621)
- [More capable agents make failures harder to detect and costlier](https://realalpha.app/s/more-capable-agents-make-failures-harder-to-detect-and/)
  MTS · Nov 8, 2026 · 4 min
  > “as models were getting better this problem was getting worse” — Ben Hylak · [1:47](https://realalpha.app/v/s_wkGlV-wr4/?t=107)
- [Agent memory failures need separate storage and retrieval metrics](https://realalpha.app/s/agent-memory-failures-need-separate-storage-and-retrieval/)
  MTS · Nov 8, 2026 · 3 min
  > “if you actually want to improve your agent you need to understand if it's from storage or retrieval” — Ben Hylak · [6:36](https://realalpha.app/v/s_wkGlV-wr4/?t=396)
- [Frontier models can be worse at adapting their tone](https://realalpha.app/s/frontier-models-can-be-worse-at-adapting-their-tone/)
  MTS · Nov 8, 2026 · 2 min
  > “i do think that frontier models are actually worse at these things than earlier models” — Ben Hylak · [14:03](https://realalpha.app/v/s_wkGlV-wr4/?t=843)

## The episode

- [Frontier AI Models Have Gotten Worse at Writing | Ben Hylak](https://realalpha.app/v/s_wkGlV-wr4/) · MTS · Nov 8, 2026