Verified Baselines Before Build: AI Discovery for a Public Manufacturer
$700M+ Publicly-Traded Manufacturer
Business Situation
A publicly-traded manufacturer routing between 15,000 and 20,000 inbound dealer requests a month through Customer Service and Design, across four systems of record, with several of them mid-migration.
The Challenge
Throughput and speed scaled with headcount rather than operational leverage, and customer conversion metrics could not be produced at all. Without a verified baseline, AI investment would be committed against estimates rather than evidence.
Why It Matters
Compose ran a five-week AI Business and Technology Discovery across both departments on a six-dimension diagnostic lens: on-site shadowing, two elapsed-time baselines verified against the client’s own SLA data, fifteen opportunities scored for value and measurability, and an architecture direction for an in-tenant v1 AI intake layer.
What Compose Delivered
- Discovery findings report across Customer Service and Design tracks
- Six-dimension diagnostic applied consistently to both departments
- Two elapsed-time KPI baselines verified against client SLA data
- Opportunity map scored on value, measurability, feasibility and TTV
- Architecture direction for an in-tenant, system-agnostic AI layer
- Metrics framework with methodology and measurement cadence
Outcome
2 elapsed-time KPIs verified against documented standards and the first measured baseline for either one.
The second KPI nests inside the first, so gains flow rather than add. Both were verified against the client’s own SLA reporting, and the development and implementation phase was costed and modeled directly against them.