Moda's agent observability weekly: Jev for sharper signals, whole-conversation analysis for long-running agents
KlausCodes · x · 2026-09-23
Agent observability startup Moda launched weekly engineering updates. Issue 1 highlights:
- Jev classifier upgrade: their homegrown classifier screening signals like tool failures and user frustration struggled with customer-specific signals due to data scarcity; with Jev they report much higher confidence and accuracy on internal benchmarks.
- Whole-conversation analysis: a single scan of long-running agent traces misses cross-context failures (e.g. an early instruction "ask before changing billing" violated 70 turns later), so Moda now reads entire conversations to surface such issues.
- Product updates: rebuilt dashboard and trace views, a public Slack agent for querying alerts directly, and import of Braintrust history to mine past traces for signals.
More from coding & agent
- User runs Opus 5.5 at max to write a ComfyUI wildcard — it did 20,000 iterations over a/an grammar — mwoody450 · 2026-09-23
- Solo creator builds AI action short with MiniMax, Tripo and Blender — shares full workflow — Spoonman915 · 2026-09-23
- Microsoft's Agensh Scales Multi-Agent Systems to 1,024 Agents Without a Central Orchestrator, Boosting Test-Pass Rate to 55% — andrew_n_carr · 2026-09-23
- Viral demo claims 'GPT-6' can drive browser Paint to draw, unverified — alexcovo_eth · 2026-09-23
- Parallel coding agents merge cleanly and silently break every test — RunAI_Coder · 2026-09-23
- Getting phone-captured text into your computer: OCR, vision models, agentic pipelines — silenceimpaired · 2026-09-23