Comply v2 - GA Deployment Plan
Team will deploy Comply v2 GA on April 7th after night-before infrastructure changes.
Internal sync to align on Comply v2 GA rollout. The team confirmed the sequence: infrastructure changes on the night of April 6th, then flipping the feature flag to GA on April 7th. Ravi reported a clean dry-run, Mike completed code review, and Nina's QA regression passed. The only concern is p99 latency above SLO under 50 concurrent requests, which the team deemed low risk due to phased rollout. Ananya will update the runbook and circulate it for final review by EOD Thursday.
How the call went
No external-side turns; all sentiment fields default to neutral.
Opening
0Close
0No turn-level sentiment was scored for this call, so only the opening, close, and overall figures above are available.
How the conversation ran?Talk share is measured from speech time in the captured portion of the transcript.
- Questions we asked
- 15
- Questions they asked
- 0
- Turns
- 39
- Longest silence
- 2s
Who was on the call
Customer
- Internal call
What happened9
The moments the model picked out, grouped by kind. Each opens onto the turns behind it.
Issue raised1?A problem was brought up on the call.
Mike reported code review complete and noted hardcoded version strings in the PCI DSS report, filed as a post-GA cleanup ticket.
Commitment3?Someone committed to doing something.
Ravi flagged cold start latency on SOC 2 report generation and committed to adding a Lambda warm-up ping to the pre-deploy checklist.
Nina committed to setting up a queue depth alert on the reporting service; Ravi agreed to work on it the next morning.
Ananya committed to updating the runbook with alert config and warm-up ping note and sending it around for final review by EOD Thursday.
Decision3?A decision was reached.
Team aligned on GA sequence: infrastructure changes night of April 6th, feature flag flip to GA morning of April 7th.
Team accepted that 50 concurrent requests is low risk for day one because the flag rollout is phased across tiers.
Team confirmed rollback via feature flag and keeping v1 in parallel for two weeks post-GA.
Risk1?A risk to the account, project, or system was surfaced.
Nina reported regression suite passed but p99 latency reached around 22 seconds at 50 concurrent requests, above the SLO target of 15 seconds.
Recap1?A summary of what was agreed.
Ravi described redundancy improvements—three processing nodes and a circuit breaker—informed by the March Detect outage.
What was promised4
| Commitment | Owner | Due | Action type | ||
|---|---|---|---|---|---|
| Add Lambda warm-up ping to pre-deploy checklist | Ravi GuptaAegisCloud | No date given | process improvement?Update or introduce internal workflows, checklists, runbooks, gates, policies, or training to change how work is done and prevent recurrence, distinct from product/system changes and static documentation. | ||
| Set up specific alert for queue depth on reporting service | Nina KowalskiAegisCloud | 2 Apr 2026said “tomorrow morning” | engineering or configuration change?Implement, deploy, fix, configure, activate, enable, migrate, backfill, or decommission a product, system, or infrastructure, including code changes and environment configuration. | ||
| Update runbook with alert config and warm-up ping note, and send around for final review | Ananya SharmaAegisCloud | 2 Apr 2026said “by EOD Thursday” | process improvement?Update or introduce internal workflows, checklists, runbooks, gates, policies, or training to change how work is done and prevent recurrence, distinct from product/system changes and static documentation. | ||
| Decommission v1 reporting service after two weeks post-GA ComplyImplied only | No owner | 21 Apr 2026said “two weeks post-GA” | engineering or configuration change?Implement, deploy, fix, configure, activate, enable, migrate, backfill, or decommission a product, system, or infrastructure, including code changes and environment configuration. |
Themes and issues4
What the call was about, and the specific issue recorded under each theme.
| Theme | As the model phrased it?The wording the model used before mapping it to the shared taxonomy. | Issue | |
|---|---|---|---|
| Product Roadmap?Calls about the product roadmap, planning, launch, release, and delivery of new capabilities. | ga launch | deployment sequence | |
| Reliability?Calls about overall system reliability, availability, resilience, capacity, and performance, excluding specific outages or incident response. | performance | report generation latency | |
| Reliability?Calls about overall system reliability, availability, resilience, capacity, and performance, excluding specific outages or incident response. | resilience | redundancy and circuit breaker | |
| Infrastructure?Calls about infrastructure, architecture, deployment, upgrades, and system configuration. | rollback management | feature flag rollback plan |
Numbers stated on this call9
Values are shown exactly as they were spoken.
| Claim | As spoken | Metric | Stated by | |
|---|---|---|---|---|
| GA deployment date | “April 7th” | date | Ananya Sharma | |
| Deploy window time | “around 10 PM Eastern” | duration | Ananya Sharma | |
| Buffer before early morning traffic | “about a three hour buffer” | duration | Ravi Gupta | |
| SOC 2 first-call latency | “like 8 seconds” | response time | Ravi Gupta | |
| Concurrent requests in load test | “50 simultaneous requests” | count | Nina Kowalski | |
| p99 latency at 50 concurrent Exceeds stated SLO target of 15 seconds. | “around 22 seconds” | response time | Nina Kowalski | |
| SLO target for p99 latency Unit seconds inferred from context. | “15” | response time | Nina Kowalski | |
| Tier one customer count | “a few hundred orgs” | account count | Ananya Sharma | |
| v1 decommission timeline | “two weeks post-GA” | date | Ravi Gupta |
How far to trust this call
Every fact above was extracted from the captured portion of the transcript only.
8 min captured of 30 min stated. Anything said in the remaining 22 min is absent from this page.
0 turns flagged low-confidence.
- Clocks unaligned(warning)
4 participant(s) speak before joining — transcript and events are on different clocks. Do NOT derive absolute timestamps by joining these two sources.
- Partial transcript(warning)
transcript covers 25.1% of a 29.5min meeting; 21.5min after the final turn is unaccounted for
The model's own note: Transcript covers only 7.9 minutes of an internal planning call; partial transcript.
Extracted 2 Aug 2026 by deepseek-v4-flash · extractor v1.0.0 · schema v1.1.0 · extended thinking on · 17,127 tokens