Tag
7 posts
1 post
Building a voice agent that takes real phone calls: HMAC-verified webhooks, a sideband transcript pump, delegation-based tool dispatch, and a refusal layer protected by cryptographic digests.
6 posts
Voice agents feel synchronous but live on a mess of tool calls, reconnects, and long-running jobs. A background job manager with dedup, bounded retries, a dead-letter queue, and a sideband control channel — plus a discovery system for assumptions you cannot verify.
How does an assistant remember you across sessions? A speaker-attributed extraction pipeline with dedup and rate limits, three isolation scopes, and a retrieval hierarchy that ranks topics, decays by recency, and never truncates silently.
How do you compare LLMs on generating structured queries? Syntax similarity is worthless. What works: executing generated queries against a live triplestore and comparing result sets — plus a spread rule most benchmarks ignore.
How to make a language model reliably drive a structured UI: force a tool call every turn, treat the final answer as a schema, feed validation errors back for self-correction, and bound the whole thing by construction.
What if a language model compiles a procedure once and never runs again? A skill compiler with hard structural gates, a zero-model interpreter that walks the graph in plain code, and a trust ladder that must be earned run by run.
When a model answers over private knowledge, the answer must never leak what it used. How to scrub IRIs, UUIDs, and query fragments across streaming chunk boundaries, and how to project provenance as an allowlist.