Speaking profile

On stage & in the community.

I give technical talks that respect the audience's intelligence and their time — turning MLOps, LLM observability, and the messy reality of production ML into something people can use.

Book me to speak Email an invitation

Saumya presenting to a seated audience beside an ML Conference banner, with a slide reading "So, what do we do?" projected on screen.

Topics

Topics I speak on.

LLM observability and evaluation
RAG systems and LLM customization for enterprise
MLOps tooling and the open source stack
ML pipelines and orchestration (MLflow, Airflow)
Data and annotation workflows

Talks & workshops

Where I've spoken.

Speaking at ML Conference, PyData, and MLOps Community events across multiple cities in Europe since 2022.

Speaker profile — ML Conference

Speaking at

Workshop
Streamlining Data Workflows with Voxel51 and Label Studio
with Thomas Waller
ML Conference · Munich2025
Designing data and annotation workflows with Voxel51 and Label Studio.
Talk
You Can't Improve What You Can't See: Making LLMs Transparent & Accountable
with Saif Addin Ellafi
ML ConferenceBerlin 2025 · Amsterdam 2026
LLM observability and evaluation — building transparency into GenAI systems in production.
Workshop
Operationalizing GenAI Agents: Deployment, Orchestration & Observability with LangGraph, LangFuse & Airflow
ML Conference · Berlin2025
A full day — eight hours — on deploying, orchestrating, and observing GenAI agents with LangGraph, LangFuse, and Airflow.
Workshop
Workshop: MLFlow, LabelStudio, FiftyOne — Harnessing Tools for Efficient Machine Learning
ML Conference · Munich2024
Talk
Reusable AI: Customizing LLMs for Diverse Business Needs
with Saif Addin Ellafi
PyData Munich · Munich2024
Customizing RAG systems with LLMs so one solution serves many products instead of handling each demand separately.
Talk
Leveraging Open Source Tools for Advanced Machine Learning Pipelines
MLOps Community Munich · Munich2024
The open source stack at each stage of the MLOps workflow.
Workshop
Managing Machine Learning Life Cycles with MLflow
ML Conference2023
MLflow and Apache Airflow for building and operating an ML pipeline.
Talk
ML Model Monitoring
with Alon Gubkin (CTO, Aporia)
PyData Munich · Munich2023
Monitoring ML models once they're live — drift, performance degradation, and what's worth instrumenting.