Hi Adam, let's dive into your data.
How can I help you today?
Review Delivery Risk
Identify at-risk deliverables and review suggested remediation steps.
Suggest Improvements
Identify roadblocks and inefficiencies, and enact improvements.
Review AI tool adoption
Understand AI tool adoption trends on your team.
Jellyfish Assistant
In February 2026 we saw a spike in defects opened. Can you figure out why?
Reasoned in 1 step

February 2026 Defect Spike — Root Cause Summary

Company-wide defects jumped ~36% from January to February (236 → 322 opened), but the spike wasn't evenly distributed. A handful of teams drove the majority of it, each for different reasons.

What the Data Shows

TeamJan → Feb DefectsChangePrimary Signal
Navy3 → 31+933%Sprint addition rate of 35x — overwhelmed with unplanned work
Blue16 → 45+181%Sprint additions tripled — reactive defect flood
Lemon4 → 11+175%KTLO allocation nearly doubled heading into February
Violet9 → 17+89%Moderate increase, resolution improving
Plum91 → 315 day cycle+245%Work stalling in development, not completing cleanly
ScarletKTLO: 1.37 → 0.18 FTENear-zero maintenance capacity alongside peak feature output

Action Items

  1. Post-mortem on Navy — A 933% spike paired with a 35x sprint addition rate points to a specific event, likely a bad release or integration. Identify it, contain it, and document what broke down.
  2. Audit Blue's February releases — Review what shipped and whether adequate testing gates were in place before it hit production.
  3. Unblock Pine and Plum pipelines — Work sitting in development for 39–52+ days is a quality time bomb. Find what's blocked and clear it before it compounds into more defects.
  4. Rebalance KTLO investment and sprint scope controls — Scarlet and Dandelion need a minimum maintenance floor (10–15% KTLO) to avoid accumulating quality debt. Navy and Blue need a sprint change policy to prevent unplanned work from flooding the team again.
Reply here
AI can make mistakes, so double-check responses.