Jan Schweiger
Hybrid AI: Next Generation Natural Language Processing
#1about 1 minute
Why 90% of AI projects fail in production
Most AI projects fail to reach production due to challenges with accuracy, data quality, and robustness in real-world scenarios.
#2about 5 minutes
How modern NLP uses Transformer models for search
Transformer models understand the full context of a sentence, enabling semantic search by converting text into vectors for comparison.
#3about 1 minute
Why pure Transformer models fail in the real world
Transformer-only models often struggle in production due to inefficiency, reliance on domain-specific training data, and a lack of robustness.
#4about 2 minutes
The strengths of classical NLP and keyword search
Classical NLP methods like BM25 keyword search are computationally efficient, require no training data, and are highly robust across different domains.
#5about 1 minute
Combining models with the hybrid AI approach
Hybrid AI combines the high accuracy of modern NLP with the efficiency and robustness of classical methods to create superior production models.
#6about 3 minutes
How to build a hybrid search engine with Vespa
Vespa is an open-source tool that simplifies building hybrid systems by allowing you to define parallel search pipelines for Transformers and BM25.
#7about 2 minutes
Analyzing the performance of a hybrid search model
The hybrid AI approach was four times faster than a pure Transformer model while maintaining high accuracy and robustness.
#8about 2 minutes
Exploring other real-world use cases for hybrid AI
Hybrid AI can be used for expert identification by building correctable knowledge graphs and for safety-critical systems like train controls.
#9about 3 minutes
Recap and recommended tools for building NLP models
A summary of how hybrid AI balances deep learning's accuracy with rule-based systems' robustness, plus recommended libraries to get started.
Related jobs
Jobs that call for the skills explored in this talk.
Matching moments
03:08 MIN
Enabling hybrid AI with an open software stack
Bringing AI Everywhere
03:31 MIN
Previewing the "AI or knockout" conference talk
From Learning to Leading: Why HR Needs a ChatGPT License
01:06 MIN
Moving beyond hype with real-world generative AI
Semantic AI: Why Embeddings Might Matter More Than LLMs
03:10 MIN
A rapid-fire look at AI tools and buzzwords
Rethinking Customer Experience in the Age of AI
02:52 MIN
Envisioning hybrid teams and 90% AI-written code
Agents for the Sake of Happiness
01:21 MIN
Combining human and AI strengths for better decisions
AI for decision-making in Tech Recruiting
02:54 MIN
Common patterns and challenges in enterprise AI adoption
WWC24 - Beyond the Hype: Real-World AI Strategies Panel
02:09 MIN
The future of translation and human-AI collaboration
Fireside Chat: Deep Learning, Deep Impact: Harnessing AI for Language Innovation
Featured Partners
Related Videos
35:16How AI Models Get Smarter
Ankit Patel
26:05Semantic AI: Why Embeddings Might Matter More Than LLMs
Christian Weyer
26:04HR ROBO SAPIENS: Decoding AI Agents and Workflow Automation for Modern Recruitment
José Kadlec
32:26Bringing the power of AI to your application.
Krzysztof Cieślak
24:24Best practices: Building Enterprise Applications that leverage GenAI
Damir
26:54Make it simple, using generative AI to accelerate learning
Duan Lightfoot
24:47How to Decipher User Uncertainty with GenAI and Vector Search
Ben Greenberg
25:14The shadows of reasoning – new design paradigms for a gen AI world
Jonas Andrulis
Related Articles
View all articles
From learning to earning
Jobs that call for the skills explored in this talk.
wabcon
Remote
DevOps
Docker
FastAPI
WebPack
+2
Neural Concept
Lausanne, Switzerland
DevOps
Continuous Integration
Nomitri
Berlin, Germany
DevOps
Gitlab
Docker
Ansible
Grafana
+6
