Title here
Summary here
Value Proposition
Executive Summary
Fred is a fully open-source agentic + RAG platform.
It is designed to bridge the gap between:
- experimentation with LLMs,
- and production-grade AI systems with governance, security, and operational constraints.
What Fred Is
Fred is the open-source technical foundation:
agentic-backendfor agents, sessions, streaming, model routing, and orchestrationknowledge-flow-backendfor ingestion, metadata, retrieval, and content lifecycle- React frontend for user-facing and admin-facing interactions
- policy-based governance primitives (models/rules, RBAC/ReBAC, observability)
Fred is designed to be:
- tool-first
- production-oriented
- modular and extensible
- deployable on-prem or cloud
Open Assets Available to Everyone
Anyone can take and use:
fredrepository (core platform)fred-deployment-factoryrepository (deployment examples and templates)
These assets already provide strong value. For real deployments, teams still need to provision and operate surrounding services (for example OpenSearch, MinIO, Keycloak, databases, monitoring, and platform run operations).
Why Fred Enables Scalable Delivery
Fred supports a clear operating model with separated responsibilities:
| Responsibility | Platform Team | Delivery & Operations Teams | Business/Product Teams |
|---|---|---|---|
| Core platform engineering and reusable enablers | Primary owner | Consumer | Contributor (optional) |
| Platform roadmap and governance patterns | Primary owner | Input + adoption | Input from use cases |
| Release management and platform quality gates | Primary owner | Consumer | Consumer |
| Deployment blueprint and runbooks | Co-owner | Co-owner | Consumer |
| Deploying a concrete instance | Support | Primary owner | Consumer |
| Operating and maintaining each deployed instance | Support | Primary owner | Consumer |
| Security accreditation execution for each environment | Guidance + artifacts | Primary owner | Consumer |
| Building domain prompts/agents/corpus configurations | Framework and best practices | Platform support | Primary owner |
| Developing domain-specific agents and workflows | Framework and reviews | Platform support | Primary owner |
This model keeps the platform team focused on reusable capabilities while allowing delivery teams to scale concrete deployments and business teams to focus on domain outcomes.
Typical Adoption Pattern
- Start from Fred and deployment assets.
- Establish a platform ownership model (governance, release, architecture).
- Deploy one or more instances with delivery/run teams.
- Let business teams configure and build domain assistants.
- Feed reusable improvements back into the platform.
Why This Matters
Without a platform-centered model, organizations often mix:
- platform engineering,
- deployment operations,
- and business solution delivery.
Fred helps establish this separation with an open, modular, and governance-ready foundation.