TeamOhana said on Nov. 10 that it has launched agent-aware workforce planning, positioning the capability around planning human workers, AI agents, or a mix of both against an approved budget. The company’s related materials also describe general headcount-planning and what-if scenario workflows. Those are the documented claims and concepts available from the sources below; they should not be read as proof of specific UI behavior, full-package availability, or guaranteed outcomes.
A practical way to use a new planning model is to test one role first, before changing a live workforce plan. In a hypothetical documentation-triage example, a team could compare one existing internal role against a proposed human-plus-agent workflow using the same anonymized or synthetic tickets. The point is not to replace people automatically, but to see whether routing quality, exception handling, and budget treatment actually hold up in a controlled scenario.
What the announced capability and documented concepts say
The announcement says TeamOhana is making agent-aware planning available so organizations can consider humans, AI agents, or combined assignments when planning against budget. Separately, TeamOhana’s planning guide and what-if pages document the broader idea of modeling scenarios, comparing staffing options, and testing assumptions before acting. Those source materials support scenario planning as a concept, but they do not by themselves prove every possible implementation detail or business result.
Confirm what is available in your own plan before designing the test. Keep the comparison separate from the approved workforce plan so the exercise does not change any live records.
A reader-ready test for one role
Use a single existing internal role in a hypothetical scenario, such as documentation triage. Keep the test narrow enough that you can compare like with like. Feed both versions the same 20 synthetic requests and record three things: how many requests are routed to the correct destination, how many minutes of human review are still needed, and how many unresolved exceptions remain at the end of the review cycle.
Then assign exactly one exception owner for the cases that do not resolve cleanly. If a request needs human judgment, make that visible rather than smoothing it away. The test should also separate a subscription or platform fee that is already counted elsewhere from any incremental cost created by the new scenario. That keeps the comparison from double-counting the same expense.
For example, suppose both workflows route 18 of 20 requests correctly, and the assisted workflow takes 12 minutes of human review instead of 20. If it misses one required escalation and the baseline misses none, it fails a zero-miss rule even though it saves eight minutes of review in this fictional test.
How to compare the two versions
- Baseline: one existing human-led documentation-triage role using the current process.
- Scenario A: the same role with an AI agent handling defined routing steps while a human reviews exceptions.
- Input set: reuse the same 20 fictional requests in both scenarios.
- Metrics: correct destinations, missed escalations, human review minutes, unresolved exceptions, and incremental cost beyond any subscription already counted elsewhere.
- Decision rule: assign one exception owner and reject any scenario that misses even a single required escalation if the planning standard requires zero misses.
Because this is a planning test, not an operational change, the team should avoid drawing conclusions about employment actions or live service redesign from the first pass. If the model looks promising, repeat the exercise with another role or a second synthetic ticket set before making broader decisions.
Source links
TeamOhana launch announcement on agent-aware workforce planning
TeamOhana headcount planning guide
TeamOhana what-if scenarios page
