Vinay Kumar, CEO and founder at Arya.ai at Cypher2017
summary
Arya.ai describes a platform with an open-source framework called Braid, a Veda workbench, GUI drag‑and‑drop model building, automated hyperparameter tuning and single‑click production that deploys hardware‑agnostic models. A deployed insurance claims model ingests 460+ input features, runs on one GPU at <0.04s per claim, achieves 95–96% precision and automates 80–85% of claim decisions, with transfer‑learning, sandboxed API composition and partnerships with Google, NVIDIA and Oracle Singapore.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 7:03Challenges in Deep Learning
- 12:05Current Solutions
- 13:38Learning Retraining
- 14:25Deep Learning
- 16:06Product Strategies
- 16:53Ecosystem
- 17:40Platform
- 18:27Advanced Data Scientist
- 20:55Sandbox Server
- 21:17Autonomous Learning
- 21:42Retraining
- 22:01Testing
- 22:55Parameters
- 23:28Product Stack
- 24:30Loss of Value
- 24:55Neural Network
- 26:17Creation modulus
- 26:30Use cases
related talks
12mSiddharth Khullar, Co-founder & CEO at Aris Machina | Slush 2025
20mAI : Solving challenges and driving Innovation across Industries
24mCognitive claims handling with AI
37mHow YugabyteDB Empowers AI-powered Fraud Intelligence
13mTransforming Insurance Through AI
09mAI-Driven Insurance Operations
31mAgentic AI
10mGenesis Final Round: Peris.ai