Magicare AI cuts nursing admissions to under a minute, slices inference costs 33% on Baseten
baseten · x · 2026-10-15
Baseten published a case study on Magicare AI, which builds an "operating system" for post-acute care admissions. Reviewing referral packets of hundreds of pages traditionally takes facilities 30-60 minutes; Magicare's agentic pipeline parses 250 data points per packet and generates admission recommendations, enabling decisions in under a minute — up to 100x faster.
Key numbers:
- 33% inference cost reduction
- 60M TPM of traffic handled, absorbing bursty real-time admission spikes
- 2 days saved on average per rate limit request
The pipeline fires hundreds to thousands of parallel requests to parse long context and cross-reference facility capabilities, demanding low latency, high uptime, and burst tolerance — all powered by Baseten inference.
More from coding & agent
- Stanford's CausalSmith automates causal inference research with Lean-verified proofs — StanfordAILab · 2026-10-15
- ASENA: a coding agent writes robot programs on the fly, no training needed — zhengyiluo · 2026-10-15
- Thorsten Ball: no PRs, no code review — agents push to main and self-verify — kieranklaassen · 2026-10-15
- The HTTP 'referrer' typo from the '90s will never be fixed — not even by AI — jimmykoppel · 2026-10-15
- Stop letting Claude or Codex pick your stack: write it in AGENTS.md and gate new dependencies — weswinder · 2026-10-15
- Dev closes all PRs: let agents patch open source via Nix overlays instead — kieranklaassen · 2026-10-15