Customer Support Chat Transcript Corpus
Large-scale customer-support conversations generated from coherent synthetic companies, products, plans, policies, account states, issues and dialogue trajectories.
Intended applications
SFT, post-training, customer-support agents, dialogue-state learning, escalation and recovery evaluation
Dataset details
- Catalogue reference
- AMS-ENT-002
- Data family
- Enterprise Corpora
- Classification
- Synthetic
- Domain
- Customer Support & Conversational AI
- Scale / access
- 4.2B exact-token production target
- Format / delivery
- Streaming JSONL shards + manifests
- Evaluation access
- Evaluation pack on request
- Current status
- Production program
Qualification before commitment
The 4.2B figure is the configured production target; final delivered token count and completion status are represented from the production manifest.
- 01Requirement fit
Lock intended use, target scale, format, metadata and acceptance criteria.
- 02Representative evaluation
Review the available sample, case, worked pack or product materials where supported.
- 03Documentation & rights
Confirm provenance, permitted use, de-identification or generation position as applicable.
- 04Commercial scope
Finalize delivery, licence or service terms, milestones and acceptance.
Listing and delivery controls
This public technical profile supports buyer discovery. It does not by itself transfer rights, guarantee the final delivery volume or replace definitive technical, licensing and commercial terms.
Review Data Listing Policy →