GH Repository · emcie-co
parlant
Build reliable customer-facing AI agents with Parlant: an interaction control harness optimized for controlled, consistent, and predictable LLM interactions.
- stars
- 18,290
- 30-day movement
- +21/day
- Related entries
- 60
- Connections
- 3
python/poetryai-agentsopenaipython/uvgenaicustomer-servicecustomer-successPythongeminillama3hacktoberfestagent-appllmpythonai-alignment
Parlant is a Python framework for building customer-facing AI agents, described as an interaction control harness that keeps LLM behavior controlled, consistent, and predictable. It works with models such as OpenAI, Gemini, and Llama 3.
Use it when you need customer-facing LLM agents whose responses stay predictable and on-script rather than free-form.
Use it to
- Build customer service chat agents
- Constrain LLM responses to business rules
- Deploy customer success assistants
- Align agent behavior with brand guidelines
- Prototype agents across OpenAI, Gemini, or Llama 3
For Developers building controlled, production customer-facing AI agents
- Role
- agent-app
- Language
- Python
- Licence
- Apache-2.0
- Forks
- 1,555
- Open issues
- 17
- Last push
- 2026-07-12
- Latest release
- v1.1.0 · 2024-12-18
topicsai-agentsllmcustomer-servicepythongenaiai-alignment