Build AI systems that process images, audio, and video—beyond language models

Language models were the opening move. The systems shaping the next phase of enterprise AI process images, audio, and video — and the question is no longer whether this works, but where classical computer vision and speech recognition still outperform LLMs, and where the crossover is already happening. This track covers both sides of that transition, grounded in use cases that are production-ready today.

Multimodal AI

Learn from Industry Leaders about:

  • Vision-language models: practical use cases and limitations in enterprise deployments
  • Computer vision with LLMs vs. classical deep learning: when to use which
  • Speech recognition, audio processing, and voice AI in production systems
  • Video understanding: current enterprise-grade capabilities and architectural patterns
  • Multimodal RAG: combining structured data, documents, images, and audio in retrieval pipelines
  • Cost and latency trade-offs in multimodal inference at scale
  • Recommender systems and predictive maintenance

Frequently Asked Questions

What will I learn in the ML & Deep Learning track?

You’ll learn to design, train, evaluate, and deploy deep learning models using proven architectural patterns. The track balances foundational machine learning principles with cutting-edge deep learning advances like transformers and transfer learning.

How are deep neural networks designed and trained?

You’ll explore how to select architectures, choose loss functions, and tune hyperparameters. Topics include optimization strategies like Adam, batch normalization, and regularization for effective model convergence.

What is supervised learning and why does generalization matter?

Supervised learning uses labeled datasets to train models. You’ll learn how to prevent overfitting, evaluate performance, and ensure your models generalize to unseen data using methods like cross-validation and dropout.

How does explainable AI (XAI) support model trust?

XAI techniques—like SHAP, LIME, and attention visualization—help developers interpret model decisions, improve transparency, and meet compliance requirements.

What role do transformers and LLMs play in deep learning?

You’ll examine the architecture behind transformer-based models, understand how pre-training works, and apply transfer learning to build domain-adapted models with fewer resources.

Track Speakers London 2026

Track Speakers San Diego 2026

San Diego's program will go live soon! Until then, please take a look at

New York's 2024 Program!

TRACK SPEAKERS MUNICH 2026

TRACK SPEAKERS NEW YORK 2026

Track Speakers Amsterdam 2026

Track Speakers Berlin 2026

Track Program London 2026

Track Highlights 2026

Track Program Munich 2026

Track Program New York 2026

Track Program Berlin 2026

Track Program Amsterdam 2026

Track Sessions London 2026

Track Sessions London 2026

View all sessions

Track Sessions Amsterdam 2026

Track Sessions Amsterdam 2026

View all sessions

Track Sessions San Diego 2026

Track Sessions San Diego 2026

View all sessions

Track Sessions MLcon Munich 2026

Track Sessions MLcon Munich 2026

View all sessions

Track Sessions MLCon New York

Track Sessions MLCon New York

View all sessions

Track Sessions MLcon Berlin 2026

Track Sessions MLcon Berlin 2026

View all sessions

Frequently Asked Questions

What will I learn in the ML & Deep Learning track?

You’ll learn to design, train, evaluate, and deploy deep learning models using proven architectural patterns. The track balances foundational machine learning principles with cutting-edge deep learning advances like transformers and transfer learning.

How are deep neural networks designed and trained?

You’ll explore how to select architectures, choose loss functions, and tune hyperparameters. Topics include optimization strategies like Adam, batch normalization, and regularization for effective model convergence

What is supervised learning and why does generalization matter?

Supervised learning uses labeled datasets to train models. You’ll learn how to prevent overfitting, evaluate performance, and ensure your models generalize to unseen data using methods like cross-validation and dropout.

How does explainable AI (XAI) support model trust?

XAI techniques—like SHAP, LIME, and attention visualization—help developers interpret model decisions, improve transparency, and meet compliance requirements.

What role do transformers and LLMs play in deep learning?

You’ll examine the architecture behind transformer-based models, understand how pre-training works, and apply transfer learning to build domain-adapted models with fewer resources.

Enjoying the content?

Get the most out of MLcon by becoming a free community member — curated resources, weekly newsletter, and member-only perks.

Weekly
Articles + tutorials

The reads you'd find if you had time

2× / mo
Live webinars

Experts you can actually ask

Monthly
Magazine + whitepapers

Deep dives worth your weekend

On-demand
Recordings + courses

Past conferences, ready when you are

Behind the Tracks

AI Agents & Agentic Workflows
Build reliable AI agents and autonomous workflows for enterprise production

MLOps & Open Source LLMs
Deploy, and operate AI models securely—from fine-tuning to production monitoring

Multimodal AI
Build AI systems that process images, audio, and video—beyond language models

AI Strategy, Organization & Governance
Scale AI responsibly—align regulation, human accountability, and organizational readiness

Advanced RAG
Master retrieval-augmented generation—from vector search to production-grade knowledge systems