Roadmap

Welcome to the Open-Xtract roadmap. This page highlights our current progress, upcoming features, and long-term goals for the project. Our mission is to make document parsing, extraction, and processing open, lightweight, and provider-agnostic, so developers can build powerful pipelines without lock-in.

---

Recently Completed

• Core extraction pipeline (JSON outputs, standardized schema)
• Provider integrations:
  • OpenAI
  • XAI
  • OpenRouter
  • TogetherAI
  • Groq
  • Bedrock
  • Anthropic
  • Gemini
• Multi-format document support (PDF, text, etc.)
• CLI interface for quick testing
• Initial documentation

---

In Progress

• Improved error handling and retries across providers
• Expanded test coverage (unit + integration)
• Wiki expansion and richer examples
• Developer feedback loop

---

Near-Term (Q4 2025)

• Streaming support across providers and pipelines
• Additional document types (e.g. Word documents, structured text formats)
• Provider routing: automatic failover, fallback, and performance-aware selection
• Developer UX improvements:
  • Installation refinements
  • Config templates
  • Notebook tutorials

---

Long-Term Vision (2026 and beyond)

• Remain lightweight and provider-agnostic by design
• Evaluation and benchmarking suite for comparing provider performance
• Open discussion on ecosystem integrations (Salesforce, Notion, Slack, etc.) — explored only if community need emerges
• Community-driven extractors and provider templates

---

How to Contribute

We welcome contributions of all sizes:

• Add new providers
• Expand document type support
• Improve docs and examples
• Suggest use cases and workflows
• File issues and share feedback

See our CONTRIBUTING.md for setup, testing, and coding guidelines.