Engagement 06
Monthly · Quoted monthly
Maintenance
Keep the system healthy after launch.
AI systems drift. Providers change APIs. Prompts rot. Usage scales. A monthly retainer is the honest cost of keeping a shipped system sharp — by the team that built it.
Post-launch support, model updates, ongoing features
- 01
Predictable monthly capacity for fixes and improvements
- 02
Faster response when models, APIs, or traffic change
- 03
Continuous iteration without reopening a full build engagement
Best for
- 01Clients who just shipped an MVP or full product with us
- 02Teams that need model / prompt upkeep without hiring a full ML ops seat
- 03Products that need steady feature iteration after go-live
Out of scope
- Large new modules that belong in a fresh fixed-scope build
- Unlimited same-day turnaround unless contracted
- Work on systems we did not build, unless separately onboarded
What you get
Stability
- Monitoring review and incident response within agreed hours
- Bug fixes and dependency updates in the retained stack
- Provider / API change adaptations
AI upkeep
- Prompt and eval updates as quality drifts
- Model version changes and cost / latency tuning
- Guardrail and rate-limit adjustments as usage grows
Iteration
- Scoped feature work within the monthly capacity
- Backlog grooming against your priorities
- Monthly summary of what shipped and what is next
Scope of work
- 01
Retainer setup
Quoted after we know the system: stack, traffic, and how much change you expect.
- Capacity band (hours or outcomes per month)
- Response / turnaround expectations
- Channels: Linear, Slack / email, and release cadence
- 02
Monthly cycle
A repeating loop of stabilize → improve → report.
- Priority intake at the start of the month
- Fixes and features within capacity
- End-of-month notes and carry-over
How it runs
I
Quote
Based on the shipped system and expected load.
II
Onboard
Access, board, and priority rules.
III
Run
Monthly capacity against your backlog.
IV
Review
Adjust capacity as the product grows.
Payment
Retainers are quoted monthly. Billing cadence is set in the retainer agreement — not the 50 / 30 / 20 build schedule used for fixed-scope engagements.
Related work
- View
LexVault
AI-native contract lifecycle platform, draft, review, redline, sign, and collaborate, with every AI answer grounded to an exact, verified line citation.
- View
AI Personalised Kids Video Platform
Personalised animated storytelling for children. Generative video pipelines, parent-safe workflows, and infrastructure tuned for viral traffic and repeat renders.
- View
AI Content Production Platform
Content and media production platform, bulk Excel-driven workflows for AI-generated images, video, and audio, with pause/resume job control.
Other engagements
Ready to scope this?
Tell us the workflow, constraints, and timeline. We’ll come back with a written plan.