Operationalizing AI in Finance: Real-Time Analytics
summary
Claims that 83% of models never reach production and that fixing post‑development plumbing unlocks value drive the session, with concrete stacks: Airflow Spark Kubeflow DVC MLflow Neptune BentoML Docker SageMaker Evidently Kafka and examples like low‑latency fraud detection, algo trading and instant credit scoring. CI/CD patterns include automated data validation, immutable audit logs and regulatory gates, monitoring uses PSI or KL divergence for drift, and deployment discusses GPU acceleration, autoscaling clusters and BI embedding with Tableau and PowerBI.
AI-written summary · about
chapters (14)
- 0:00Introduction and Speaker Background
- 0:42Agenda Overview
- 1:13ML Ops Maturity Gap in Financial Services
- 2:59Stages of the Financial ML Lifecycle
- 4:14Tech Stack for ML Ops in FinTech
- 7:00Feature Engineering in Financial Data
- 7:56CI/CD for Financial ML Models
- 8:43Real-Time Interface Architecture
- 9:41Compliance and Governance in FinTech
- 10:21Monitoring Financial ML in Production
- 11:59Drift Management in Financial Models
- 12:39Integrating ML with BI Tools in Finance
- 13:46NLP-Driven Financial Analytics
- 15:04Key Takeaways and Conclusion
related talks
26mEngineering Excellence: Platforms and Architectures for Real-Time AI Analytics
15mScaling AI-Driven Financial Services: Ethical ML at Scale
08mConsiderations for producing the best AI and Machine Learning models
54mArtificial Intelligence seen from the software development lifecycle perspective
32mMaking Models Work: How an optimized ML lifecycle can drive better ROI
41mInfrastructure for ML Applications
56mData, AI, Hype and Fear - Recipe for Disaster?
26mThe Hidden Foundation: How Effective Data Management Determines AI System Success