LanternOpsField Guide · Managed AgentOpsManaged AI Teams
LanternOps Book a diagnosis →
§ 0The retainer

AI Teams are not set and forget.

Staff change. Software changes. Vendors change. Workflows change. A deployed AI Team needs monitoring, correction, and improvement, the same way any part of your operation does. That is the work we keep doing after launch.

01
§ 1The operating view

You can see exactly what the AI Team is doing.

Work handled, approvals waiting, edit rates trending down, stuck points surfaced, and connector health at a glance. Managed AgentOps is active, hands-on operation with a view you can read.

Fig. 1 — The operating dashboardPlate I
Managed AgentOps example data Operating
Items handled 1,284 this month
Hours returned 96 to senior staff
Awaiting approval 7 in queue
Edit rate 6% down from 11%
Edit rate trend lower is better
Stuck points
  • Ambiguous matter type on 3 intakes
  • Billing connector token expired
  • New client template not yet mapped
Connector health
  • Email
  • CRM
  • Docs
  • Billing
  • Calendar
§ 2What we actually do

What we actually do on the retainer.

The retainer is a list of jobs, performed on a schedule, with a report you can read. No black box.

  • Monitor agent performance
  • Maintain secure connectors
  • Update prompts and instructions
  • Review stuck points
  • Tune evaluations
  • Analyze staff corrections
  • Manage approval queues
  • Handle tool and vendor changes
  • Produce monthly performance reports
  • Identify the next AI Team opportunity
03
§ 3The retainer tiers

Priced to the size of the operation.

Every retainer starts from one production AI Team, proven first in a three-week pilot. You move up a tier only when the work earns it.

Starter

$5k /month

One AI Team, light monitoring, monthly review.

Firms running their first supervised team.

Most common

Growth

$10k /month

Multiple AI Teams, weekly optimization, minor builds.

Firms expanding AI across functions.

Operator

$15k–25k /month

Broader AI operations, deeper reporting, more build capacity.

Firms running AI across departments.

Not run a pilot yet? Start with the three-week pilot sprint →

§ 4The owner you hire

Every AI system
needs an owner.

Most AI projects fail because nobody owns the agent after the demo. We are the owner you hire on purpose.