From Bias to Balance: TANGO’s Hybrid-Intelligence Platform
summary
Modular Python platform that serves models via a multi‑purpose REST API and an explainability layer that produces audience‑specific explanations; supports runtime feedback through sessioned invocations so a later prediction reads prior feedback. Pilots include maternal‑care triage, credit allocation and pre‑operative aorta segmentation where the model finds the full aorta unlike commercial alternatives. Everything is released open source. hybrid learning API audience‑specific explainers aorta segmentation outperforms commercial models Apache V2, 25‑partner consortium
AI-written summary · about
chapters (5)
- 0:00Introduction to TANGO project
- 2:57TANGO’s three pillars of trustworthy AI
- 6:49Real-world use cases and demos
- 10:08Live demo of TANGO platform
- 12:13Q&A on clinical feedback process
related talks
46mCan you trust your AI?
12mTrust and transparency of AI for the enterprise
03mDBTB INT Sai Moturu r
48mMastering the AI Shift: Adaptive Strategies for Software Evolution
26mWelcome to the AI Jungle! Now What?
39mTowards Useful, Reliable, and Scalable AI Agents
39mKeynote 3: How did you know? Explaining Black Box Model Predictions in Python
01mintro video