ENGINEERING2026-05-21
Model-agnostic LLM gateway routing for self-hosted deployments
Our engineering arm rolled out gateway routing so field and legal AI systems can switch models without touching application code.
DuneDive Software Development rolled out model-agnostic LLM gateway routing across our self-hosted deployments. Applications talk to one gateway; the gateway decides which model serves the request: Cloud or local, large or small.
Switching providers is now a configuration change, not a code change. The same routing layer powers our field systems and our legal & compliance platform work.
For customers, this means no vendor lock-in and a clean migration path as the model landscape shifts.