Skip to content

Maintenance & Managed AI

Keep it running and improving: monitoring, incidents, prompt, model and cost optimization, and new features.

Book a 5-minute growth call

The problem

Software degrades without attention: dependencies age, providers change models, costs creep, small bugs become support tickets. Managed AI adds prompts that drift and evaluation sets that go stale. A standing team keeps the product healthy and keeps shipping the improvements that users ask for.

What we build

  • Monitoring, uptime and incident response with clear response times
  • Prompt, model and cost optimization with a monthly quality report
  • A feature backlog delivered in a predictable monthly rhythm

Maintenance, Uptime, AI monitoring, Prompt optimization, Model upgrades, Cost optimization

How it works

  1. 01

    Onboard

    Code, infrastructure and access reviewed; monitoring and a backlog set up in the first two weeks.

  2. 02

    Operate

    Uptime, incidents, updates and security patches handled inside agreed response times.

  3. 03

    Improve

    Monthly prompt, model and cost review; features shipped from the backlog.

  4. 04

    Report

    A monthly report on uptime, incidents, spend and what shipped.

Work that proves it

  • Phare home page: uptime monitoring, alerts, analytics and status pages on one platform

    Phare

    Monitoring, incident management and status pages

    Designed and developed for the client: monitoring checks, incident flow and public status pages.

  • Webhook Relay home page: forwarding webhooks anywhere, beside a wireframe globe of connections

    Webhook Relay

    Webhook and API infrastructure with SOC 2 Type II, SSO and audit logs

    Designed and developed for the client: webhook forwarding and replay, tunnels, and the audit and access-control surface.

What you get

  • Onboarding review and backlog

  • Monitoring and incident process

  • Monthly quality and cost report

  • Security and dependency updates

  • Shipped features from the backlog

  • Updated documentation

Questions

What's included in a maintenance engagement?
An onboarding review of code, infrastructure and access in the first two weeks, with monitoring and a backlog set up. After that: uptime and incident handling inside agreed response times, security and dependency updates, a monthly prompt, model and cost review, and features shipped from the backlog. The monthly report covers uptime, incidents, spend and what shipped.
How long does onboarding take?
Onboarding takes about two weeks, and the agreement runs month to month after that. How much lands in the first month depends on what monitoring already exists and how current the dependencies are: a product nobody has updated in a year spends its first month getting current before the backlog moves.
What do you need from us?
Access to the repository, the infrastructure and the monitoring, plus one person who can prioritize the backlog each month. For incidents we need to agree what counts as urgent and who we may call at two in the morning, before the night it matters.
Who owns the code and the data?
You do, throughout. We work in your repository and your accounts, and the monthly report and the runbook stay with you. If you end the agreement there is nothing to hand back except access, and we write an exit note covering what is due for attention in the next quarter.
What response times do you commit to?
Response times are agreed per plan and written into the agreement, and they separate an outage from a bug from a feature request — a site that is down and a misaligned button do not belong in the same queue. We publish how we did against them in the monthly report, which is the part that keeps the commitment honest.

Not sure where to start?

Describe the problem in a few lines. You get a straight answer on what we would build and how long it takes.

Book a 5-minute growth call
Book a 5-minute growth call