June 29 – July 2, 2026 · San Francisco, CA · imported from ai.engineer's public schedule feed

AI Engineer World's Fair 2026 — unofficial import demo

Unofficial demo. This programme was imported from the AI Engineer World's Fair's own public schedule feed to show vibeboard at real conference scale. Not affiliated with, or endorsed by, the organisers.

All sessions
AI in HealthcareSession

Shipping AI to a Million Patients Without an A/B Test

Jared Joselowitz

When
Thursday, July 211:40 AM – 12:00 PM · 20 min
Where
Track 7San Francisco, CA · imported from ai.engineer's public schedule feed
Google Calendar

About this session

You can't A/B test on patients. You can't unsend a phone call. The model card won't save you at the post-incident review. Most AI eng playbooks assume the opposite. Ship to 5%, watch the dashboard, roll back if it goes wrong. None of it survives regulated deployment, which is now coming for fintech, legal, and government too. So the engineering has to move: into hazard analysis, simulated populations, asymmetric evaluation, and audit trails treated as the deliverable. The trail is the product. I'll show you what changes when rollback isn't an option. How Ufonia ships Dora, an AI voice agent now making clinical follow-up calls on the NHS and across US health systems, using a hazard-driven simulation rig (MATRIX) and a prompt-optimisation flywheel that surface failures and conform the same base system to each clinical niche, all of it pinned to an audit trail. And the cheap version of all this, for any team whose users can't be the test population.

Speaker

Jared Joselowitz
Jared Joselowitz

AI Research Engineer, Ufonia

Jared Joselowitz is the Lead AI Research Engineer at Ufonia, where Dora (an AI voice agent) makes clinical follow-up calls on the NHS and across US health systems; over 200,000 patient calls delivered, with signed contracts to scale past a million. He builds the evaluation and hazard-analysis stack for clinical voice AI: multi-agent simulation, prompt-optimisation pipelines, and the audit infrastructure that has to hold up when there's a patient on the other end of the call. His research on clinical AI safety and evaluation has been published at ACL, COLM and IWSDS, most recently an LLM judge that matches clinician safety assessments of speech-recognition errors. Originally from Johannesburg, South Africa, Jared studied electrical engineer before completing an MSc in Applied Machine Learning at Imperial College London, where his thesis used inverse reinforcement learning to recover the implicit reward models of RLHF-trained LLMs.

More in AI in Healthcare