Back to directory

BISHENG

An open-source, enterprise-focused LLM application DevOps platform combining a visual generative-AI workflow builder, retrieval-augmented generation, agents, model management, evaluation and fine-tuning.

BISHENG

BISHENG is an open-source LLM application DevOps platform focused on enterprise scenarios, already used by a number of large organizations. It combines a general-purpose agent, built through an internal 'Agent Guidance Language' that embeds domain experts' preferences and business logic, with a workflow engine that runs generative-AI pipelines, retrieval-augmented generation, agents, loops, parallel branches, batch processing and conditional logic inside one visual flowchart builder, rather than separate chatbot and workflow modules for different tasks.

The workflow builder supports human intervention during a run, including in multi-turn conversations, and represents constructs such as loops, parallelism and batch processing as ordinary flowchart shapes instead of specialized components. Beyond the workflow engine, the platform covers unified model management, evaluation, supervised fine-tuning, dataset management, enterprise system administration and observability.

It targets enterprise use cases such as document review, fixed-layout report generation, multi-agent collaboration, policy comparison, customer-service assistance, meeting-minute generation, resume screening and unstructured-data governance.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.