Software and internet
From incident review to renewal brief
Agents gather the facts and draft. People make every commitment to customers.
| A | B | C | D | E | |
|---|---|---|---|---|---|
| 1 | # | Question | Brightwave response | Source | Status |
| 2 | 1.1 | Is there a named information security lead? | Yes, the Head of Security owns the program | Security org chart 1.1 | Approved |
| 3 | 2.3 | Do staff complete security training every year? | Yes, yearly and at onboarding | People security policy 2.3 | Approved |
| 4 | 4.2 | Is there a written incident response plan? | Yes, tested once a year | Incident response policy 1.2 | Approved |
| 5 | 6.4 | Are customers told about material changes in advance? | Yes, in writing ahead of the change | Library · Sep 2024 | Update |
| 6 | 9.6 | How fast are third-party vulnerabilities fixed? | — | No approved answer | Write |
186 questions · 171 answered · 15 to review
- Stale answers
- 9
- No approved answer
- 6

Agents take the paperwork beyond the code
- Outages that shake customer trust
- Security reviews that stall deals
- Churn spotted only at renewal
- 01
Build and release
Write and review code and ship on a release cadence
Change records are scattered across tickets and code reviews
Owner · Engineering and SRE
- 02
Security review
Answer customer security questionnaires and share architecture and policy documents
The same questions get answered deal after deal, and security becomes the bottleneck
Agents takeFills questionnaires from approved answers and flags new ones
Owner · Security and compliance
- 03
Onboarding
Configure, migrate data and train the customer's admins
Launch to-dos live in chat threads, and customer-side progress needs chasing
Agents takeTracks the launch checklist and drafts weekly updates
Owner · Customer success
- 04
Incidents
Run on-call, handle alerts and incidents, and update customers
Postmortems mean rebuilding the timeline from chat, alerts and deploys
Agents takeRebuilds the timeline and drafts the postmortem and customer note
Owner · Engineering and SRE
- 05
Renewal and expansion
Track usage and health and prepare renewals and upsells
Churn signals are spread across usage, tickets and contact changes
Agents takePulls the signals together and drafts the renewal brief early
Owner · Customer success
Three teams, one organization context
These teams and flows are illustrative, not a specific customer.
Work agents take on
Last night's payment webhook incident is closed. Rebuild the timeline from the incident channel, alerts and deploy log, draft the postmortem, and draft a note for affected customers. I'll review before anything goes out.
- Alerts and deploy log (MCP · read-only)
- Lark · Incident channel
- Postmortem template and past reviews
- Action item tracker
Every step is in the run trace

Where people decide
Agents prepare the work. Key decisions stay with people.
- 01
Incident commanders decide root cause, production changes and what customers are told
- 02
The security lead approves every security answer that goes to a customer
- 03
CSMs and the sales lead confirm renewal pricing, discounts and customer commitments
Start with one pilot
Try one workflow first, measure it, then roll it out.
Security questionnaire fill
The answer library already exists, every answer traces to a source, and the security lead reviews each one
Before you start
- 10 questionnaires from the past year with the final responses
- The current answer library and policies, with version dates
- The security lead and one sales engineer to review results
Measure first
- Days from questionnaire received to response sent
- Share of answers the security lead rewrites
- Answers whose source conflicts with current policy
When to stop
- Pause and fix the library if an answer without a source, or one that conflicts with policy, reaches the review draft



















