We were the people behind an AI builder's support desk.
FusionCX exists to run that operation properly for AI-native software.
Story
How we got here
We started inside an AI application builder, resolving what the model could not, verifying its diagnoses, and quietly recording what went wrong in spreadsheets nobody had asked for.
When the AI team asked whether we had any real failure examples, we had thousands. The support desk had been the best-labelled dataset in the company all along.
FusionCX exists to run that operation properly for AI-native software: certified engineers under the client's brand, commercials tied to resolution and expansion, and every ticket returned as evaluation data, kits and tests.
Principles
Principles
- 01
Publish the SLA, then attain it.
Targets are in the contract and on the dashboard.
- 02
Be invisible to the customer.
The client's brand is the only brand in the customer's experience.
- 03
Return the data, never keep it.
Work product vests in the client on creation.
- 04
Certify before contact.
No exceptions for volume or urgency.
- 05
Prefer the exact number to the strong adjective.
Every claim carries its evidence.
Team
Team
No stock photography. Roles, and one line of background each.
4
Engagement managers
Support and CS operations leads from AI-native software and enterprise SaaS.
9
Team leads
Certified on client platforms; each has built and deployed on the products they support.
1
Data quality lead
Annotation programs and inter-annotator calibration for evaluation and RLHF datasets.
1
Security and compliance lead
DPAs, sub-processor governance, access reviews and client audits.
Locations
Locations
Registered in India. Shifts are staffed from the jurisdictions disclosed in the Trust Centre. Clients in North America, India, Europe and APAC.
No commitment either way
Request the operating plan
Squads, shifts, SLAs and pricing applied to your numbers, in five business days.