JILL | Joint Interactive Learning Loop | Customer Simulator
JILL (Joint Interactive Learning Loop) is a customer service training module developed by Sunshyne Labs. Purpose JILL is designed to provide a controlled, closed-loop environment where customer service representatives can practice interactions (colloquially known in the industry as "mock calls") to help them develop better communication skills and rapport building—all without affecting real-world operational metrics on the production floor during the nesting period. Its broader and more ambitious objective, outlined in the artifact, explores how meaningful exchanges of best practices between humans and contemporary generative analogues can support the operational development of both across time. Current Version Scope This release documents an initial version of the JILL artifact and its accompanying configuration files. The included materials in this version provide the foundation from which further development of future iterations are built. The current version of this artifact outlines Yeong-Ja Kim, a simulation of an elderly customer from South Korea. JILL encompasses four customer simulations in total. Each simulation has their own persona infusion vectors, internal world models and user-inducible value cues. Future versions shall introduce the architectures for the other three simulations. Succeeding iterations may also refine the initial simulation's logic, persona, documentation, and supporting frameworks, whilst preserving the provenance of its earlier releases. Additional Information SHA-256: 4ebbeec088e4f2d1ac052979e4129d866b345d860b66661cd1ef252d9da85829 Copyright © 2026 Sunshyne Labs. The MIT License governs the permissions granted for all materials in this release. All simulations are hosted by ElevenLabs. This project is not sponsored by or affiliated with this platform, nor does the inclusion of this platform imply its endorsement of JILL or Sunshyne Labs.
Authors
- Sunshyne Labs
- Jerome Juliano
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-10-28
- DOI
- https://doi.org/10.5281/zenodo.23663473
- Primary Topic
- AI in Service Interactions
- Type
- article
- Field-Weighted Citation Impact
- 0.00