Podcast
MLOps in Finance: Regulated Deployment, CI/CD and Model Governance
Open original DataTalks.Club episode
MLOps in Finance: Regulated Deployment, CI/CD and Model Governance
Original Episode
Use these links for the canonical episode and media sources.
- Open the original DataTalks.Club podcast page
- Watch on YouTube
- Listen on Spotify
- Listen on Apple Podcasts
Episode Overview
How do you deploy machine learning in heavily regulated finance environments while keeping CI/CD pipelines, model governance, and operational risk under control? In this episode Nemanja Radojkovic—an electrical engineer turned data scientist and MLOps practitioner who moved from Belgrade to Leuven—walks through real-world constraints and pragmatic solutions for MLOps in finance.
People
Use these links to connect the episode to guest notes.
Chapter Summary
Use these checkpoints to decide whether to open the source transcript.
- 0:00 - Episode Introduction
- 1:35 - Guest Introduction: Nemanja’s journey from Belgrade to ML Ops in Europe
- 2:52 - Guest Background: Electrical engineering, PhD experience, and early career
- 8:18 - Early Data Roles: PhD, Deloitte, and first paid Python work
- 10:35 - Finance Use Cases: Compliance, AML, fraud, and smart automation (document
- 14:57 - Role Overview: ML engineering / ML Ops responsibilities in finance (CI/CD,
- 18:52 - Regulatory & Legacy Constraints: Slow change, legacy systems, and governance
- 22:25 - DevOps Governance: Release management, approvals, and building trust
- 23:39 - Integrating ML with DevOps: Adapting ML workflows to existing corporate processes
- 27:51 - On-Premises Infrastructure: Hadoop, OpenShift, hardware requests, and platform
- 31:02 - ML Ops on a Shoestring: Prioritization and minimal viable ML Ops strategy
- 31:57 - Minimal ML Ops Components: Dev/test/prod environments, monitoring, model
- 35:57 - Tactical Solutions: Using S3 and simple approaches as interim model registry/data
- 38:48 - Project Approach: Prototyping, Agile limits for ML, and iterative groundwork
- 41:14 - Team Structure: Multiple data scientists per ML engineer and standardized
- 43:39 - Platform & Reuse: Internal libraries, FastAPI framework, and maintaining
- 45:04 - Skills for ML Engineers: Python, Linux, networking, cloud basics, and stakeholder
- 48:55 - Career Transition Challenges: Moving from electrical engineering and sales
- 52:51 - Beginner Tech Stack: Python, SQL, Pandas/Polars, cloud basics, and job-market
- 56:19 - Learn by Building: End-to-end projects, web apps, and scraping job postings