An HR, operations or knowledge-service owner with an approved tool and working prototype, but little time to test and operate it.
Bring an approved existing tool, a working internal knowledge prototype, one usable source corpus and existing tests or prompts. Confirm source ownership, a technical delivery owner and the feasibility of the agreed improvement before quoting.
We test an existing internal knowledge prototype and its sources, prepare the operating rules, support authorized first use, review issues and coordinate one feasible tested improvement with a release and rollback record.
Your client owner authorizes launch, refinement or stopping. Named technical staff implement changes within the agreed scope; users provide real-use feedback and reviewers handle exceptions.
Week 1: Accept the prototype, sources and technical owner; agree baseline effort, up to 25 test questions and release criteria. Timing is confirmed after inputs and access are accepted.
Documented tests, sources and exceptions; a recorded client launch/refine/stop decision; and, where authorized and viable, four operating checks, issue disposition and the tested improvement/handover.
Only when prototype readiness, workflow boundaries, source ownership, tool approval or technical feasibility cannot be accepted. A separately priced readiness stage precedes client-authorized launch and operation; it is not a compulsory new diagnostic for usable work.
There is no working prototype, approved tool, usable corpus or named technical owner; required checks fail; or the buyer expects automated employment decisions, a new platform or custom integration within the pilot.
New platforms, custom integration and enterprise governance are separate. Human-reviewed internal knowledge support; no accuracy or savings guarantee. Failed criteria pause launch.
The accepted prototype, tool and source corpus; technical and source owners; scored release criteria; exception and human-review routes; client authorization; operating quality, issue disposition and a feasible improvement or refine/stop decision.
Illustrative example of the record we use.
The bundles are new pilots. Related case stories describe earlier Ignite work, not this bundle.
Test, Operate and Review. Agree the use case, sources and release checks; Launch the approved controlled-use cycle; Review quality and complete a bounded improvement.
One existing prototype/tool set, one source corpus and up to five users. Up to 25 test questions scored initially and once again (50 outputs); two 60-minute practice/review sessions; four weekly checks; up to ten issue triages; one feasible prompt, source or configuration improvement with tests and a release/rollback record.
Scot: Lead tests, evidence checks, operating detail and release coordination; confirm a named technical delivery owner for the agreed release.
Josh: Lead sponsor priorities, review responsibilities and launch/continue/stop decisions through existing governance.
Owner: 6-8 hours; each user: up to 2 session hours plus real-use feedback; sponsor: about 90 minutes. IT and subject-reviewer time is separately estimated and costed.
Proposed stage-priced fixed scope: readiness review, then authorized launch and operation. Confirm technical delivery owner, improvement feasibility and agreed scope; optional four-week renewal has explicit checks, issues and change caps.
Additional populations, participants, roles, sources, meetings, workstreams or review periods are agreed as separate scope where relevant.
Proposed pilot scope. We agree the remaining work, usable inputs, responsibilities and fees before starting. Timing depends on accepted inputs and the required people being available. Your team retains approvals and its assigned business actions.
For HR teams that want AI-supported speed without unsafe interpretation or unreviewed employee insight.
Scope, responsibilities and what happens next.
