SentioGate FinOps & LLM Router

A BYOK AI router and FinOps gateway that lets an assistant audit token waste, check tool-call safety, optimize payloads, and pick cost-optimal models.

What you can do

• Audit past LLM request logs to find cost waste, context bloat, prompt-cache opportunities and achievable savings
• Fetch FinOps metrics: spend by provider, prompt-cache savings, active sessions and remaining pool quota
• Recommend a model and provider fallback chain for a coding or agent task from live pricing and latency
• Pre-flight check tool calls, terminal commands or conversation history against security rules such as credential leakage, destructive commands and runaway loops
• Preview context compaction and tool-output trimming on message arrays without running inference
• Route a chat completion sub-task through the gateway or enqueue it on a background batch queue

When to use it

Useful when you want to cut LLM costs, compare models for a task, or add a safety check before running tool calls and shell commands. Also fits teams tracking BYOK spend and quota across providers.

Requirements

Runs as a remote endpoint or as a local npm package over stdio. An optional SENTIOGATE_API_KEY is needed for BYOK Vault and FinOps ledger access; network access to the SentioGate service is required.