Each 4 hrs. Spread out in 2 hr chunks.

Courses come with hands-on examples for your team to explore and take home. Courses are very conversational, workshop style, small group, focused on your team, and open to discussion. Bring questions, be ready to “stump the chump”! Access to Doug to ask questions indefinitely.

Course List

Traction on Search Relevance at you Organization

Get your team into a cycle of continuous search relevance improvement — in a way that stakeholders “feel”. How can you organize yourself as a search team to find trouble areas, plan wins, measure, and demonstrate progress?

I teach cross-functional development practices that have worked for me and that you can adapt to your organization.

In this class, you’ll learn to:

• Put a number on search quality
• Find the biggest gaps in search quality
• Pick projects that have biggest impact to those metrics
• Distinguish between “tuning” style work and larger “big swing” projects
• Turn non-technical colleagues into search allies, not search antagonists
• Sanity check improvements before launching beyond metrics
• Improve team morale by shipping and winning

Lexical + Vector → Hybrid Search

Lexical search comes from a world of precise text matching, inverted indices, and carefully managed synonyms. Vector search, however, learns the statistical properties of language - clumping together similar concepts, loosely and noisily. To build a modern search application we must learn to combine these into systems into a single holistic system.

This class covers the art of constructing a hybrid search ranking function to target a search use-case at scale.

In this course you’ll learn:

• How to measure the impact of your changes on search system
• When to prefer a lexical approach vs a vector approach to a problem
• Basic aspects of lexical and vector search data structures - their strengths and weaknesses
• How to get started with Reciprocal Rank Fusion and Interleaving
• Using lexical-friendly vector search to retrieve relevant vector candidates
• The different flavors of “semantic search” - traditional knowledge graphs, taxonomies vs embeddings

Generative AI Augmented Retrieval

We have for years manually curated synonyms, built custom rules to understand search queries. We’ve had experts build rules and taxonomies. Now it seems Large Language Models might be able to do this for us! Right? Generative-AI Augmented Retrieval (GAR) teaches an old search engine new tricks. Many teams have extensive lexical infrastructure in place

In this course you learn to take advantage of Generative AI to improve the pace of search development. You’ll learn to:

• Use LLMs to improve query understanding
• Build taxonomy and semantic search without a vector database
• Learn to classify queries and content into a shared vocabulary
• Leverage user data as examples to generative AI to calibrate to your users behavior patterns.
• Avoid pitfalls of the LLM’s lack of knowledge of your specific search app
• How to prompt an LLM for synonyms and managed vocabularies

Pricing

Each course is $12,000. Invite up to 12 teammates.

Payment up front before course delivery. Payment options:

• Option 1 - pay with stripe
• Option 2 - I can invoice you electronically

More about Doug

We haven’t met? Learn about me here