OpenAI GPT-5.5 + Codex, now available and fully-governed in Databricks
Summary
Databricks now natively supports OpenAI's GPT-5.5, a frontier model designed for agentic enterprise work, complex document reasoning, and long-horizon coding agents. This integration allows customers to apply GPT-5.5 to their enterprise data with end-to-end governance via Unity AI Gateway. The Unity AI Gateway provides centralized security, cost controls, and observability for all GPT-5.5 and Codex usage, including permissions, rate limits, PII detection, prompt injection blocking, content safety enforcement, automatic failover, and comprehensive logging. This partnership enables enhanced coding workflows, data-grounded agents, deeper document intelligence pipelines, and natural language analytics through Genie, available across AWS, Azure, and GCP.
Key takeaway
For CTOs and VPs of Engineering evaluating enterprise AI solutions, the native integration of GPT-5.5 on Databricks with Unity AI Gateway offers a compelling path to deploy advanced LLMs securely and cost-effectively. You can leverage this to build governed coding agents, enhance document processing, and enable natural language data analytics, ensuring compliance and control over your AI initiatives from day one.
Key insights
Databricks integrates OpenAI's GPT-5.5 and Codex with Unity AI Gateway for governed enterprise AI applications.
Principles
- Centralized governance is crucial for enterprise AI.
- AI models enhance natural language data interaction.
- Robust security and cost controls are essential.
Method
Databricks Unity AI Gateway governs GPT-5.5 and Codex by providing a single control plane for permissions, rate limits, guardrails (PII, prompt injection), auditability, automatic failover, and comprehensive observability.
In practice
- Use Genie for natural language data queries.
- Build custom agents with Agent Bricks.
- Automate document intelligence with Lakeflow Spark.
Topics
- OpenAI GPT-5.5
- Databricks
- Unity AI Gateway
- Codex
- Enterprise AI
Best for: CTO, VP of Engineering/Data, Machine Learning Engineer, AI Engineer, MLOps Engineer, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Databricks.