AI Engineer · Video summary
Full Workshop: Build Your Own Deep Research Agents - Louis-François Bouchard, Paul Iusztin, Samridhi
Build deep research agents and writing workflows to automate content creation, focusing on structured prompts, evaluator-optimizer loops, and observability for production-ready AI systems.
Watch on YouTube ↗Key takeaways
- 🎯Distinguish between workflows (sequential, predictable tasks) and agents (autonomous, reactive, planning capabilities) to select the appropriate system architecture for a given problem.08:12
- 💡The 'autonomy slider' helps choose the right system: simple prompting for known tasks, workflows for predictable sequences, and agents for dynamic, reactive decision-making.10:45
- 🚀Map your current AI tasks or desired automated processes to the autonomy slider (simple prompting, workflow, or agent) to determine the most efficient system architecture.10:45
- ✍️Control AI content generation by structuring user input into guidelines, using writing profiles for style and tone, and providing high-quality, representative few-shot examples.11:14
- 📝Develop structured guidelines and writing profiles (style, tone, banned words) to control AI content generation, treating them as critical inputs rather than mere topics.11:14
- 🔁Implement an evaluator-optimizer loop with separate writer and reviewer LLMs to iteratively refine content, reduce bias, and ensure adherence to guidelines, research, and profiles.13:17
Chapters
- 0:14👋 Workshop Introduction
- 0:21📝 Understanding AI Content Issues
- 0:48🔍 The Need for Deep Research Agents
- 1:51📚 Building a Technical Writing System
- 4:04👨💻 Meet the Team
- 5:30🔧 AI Engineering Challenges
- 7:31⚙️ Defining Workflows
- 10:00🤖 When to Use Agents
- 11:27📈 Real-World Use Cases
- 12:53🔄 Context Management
- 17:46💡 Deep Research System Overview
- 21:54🛠️ Building the Deep Research System
- 25:00💻 Code Walkthrough
- 31:25🎥 Analyzing YouTube Content
- 35:00📊 Final Thoughts and Q&A
The digest
High-Level Overview
This workshop comprehensively presents the design, development, and deployment of deep research agents and AI-augmented technical writing workflows to automate complex knowledge synthesis and content creation tasks. Aimed at AI engineers, technical writers, and developers, the session covers foundational concepts, architectural decisions, practical implementations, code demonstrations, and evaluation strategies. The content showcases how to combine agentic flexibility with constrained deterministic workflows to produce high-quality, verifiable, and context-aware AI-generated outputs, particularly focusing on producing technical articles and LinkedIn posts that avoid typical AI “slop” and hallucinations.
Keep reading in The Digest
The full 7-minute article, 20 more sections, and every new episode from AI Engineer and the creators you follow.
- Structured Deep Dive
- 1. Motivation and Problem Context for AI Engineering ([00:14–05:01])
- 2. Defining Workflows and Agents in AI Systems ([06:09–19:04])
- 3. Multi-Agent Systems and Context Management ([19:04–21:54])
- 4. What is a Deep Research System? ([21:54–26:14])
- 5. Architectural Decisions: Splitting Research and Writing Agents ([26:14–29:42])
- 6. Implementation Details and Tools ([29:42–35:40])
- 7. Deep Research Agent in Action ([34:46–56:18])
- 8. Transition to Writing Workflow: LinkedIn Post Generation ([01:10:35–01:30:48])
- 9. Observability and Monitoring ([01:32:59–01:36:56])
- 10. AI Evaluation (Evals) and Quality Assurance ([01:37:00–01:49:32])
- 11. Final Recommendations and Resources ([01:49:32–end])
- Key Frameworks and Mental Models
- Autonomy Slider in AI Engineering
- Multi-Agent System Delegation
- Evaluator-Optimizer Loop for Content Generation
- MCP (Modular Capabilities Protocol)
- Practical Implications & Takeaways
- Common Mistakes Addressed
- Action-Oriented Summary
More from AI Engineer
Training an LLM from Scratch, Locally — Angelos Perivolaropoulos, ElevenLabs
In this hands-on workshop, Angelos Perivolaropoulos from ElevenLabs guides participants through the process of training an LLM from scratch using basic libraries like Torch, emphasizing the importance of tokenization and model architecture in building effective language models.
Skill Issue: How We Used AI to Make Agents Actually Good at Supabase — Pedro Rodrigues, Supabase
In this workshop, AI Engineer Pedro Rodrigues discusses how to enhance agent performance at Supabase using skills—structured workflows that improve agent interaction and efficiency. He shares insights on creating, testing, and implementing these skills, highlighting their significance in the development of agent-friendly applications.
Ralph Loops: Build Dumb AI Loops That Ship — Chris Parsons, Cherrypick
In this workshop, AI Engineer Chris Parsons introduces Ralph Loops, a method for building simple AI workflows that enhance productivity by iteratively refining tasks, using real coding examples like a Pomodoro timer. Participants learn to implement these loops practically, emphasizing the importance of context and feedback in AI-assisted coding.
Mergeable by default: Building the context engine to save time and tokens — Peter Werry, Unblocked
In this video, Peter Werry from Unblocked discusses the importance of a context engine for AI agents, emphasizing how it optimizes task execution by providing relevant context and reducing cognitive load for users. He debunks common myths surrounding context engines and illustrates their effectiveness in enhancing productivity and understanding within organizations.
Context Is the New Code — Patrick Debois, Tessl
In this talk, Patrick Debois presents the concept that context is the new code, emphasizing the importance of creating, testing, and distributing context in the software development lifecycle, particularly when using AI coding agents. He advocates for a systematic approach to context development that involves generating reusable prompts, evaluating their effectiveness, and leveraging feedback to improve overall coding practices.
Mastering AI Pricing: Flexible & Agile Monetization — Mayank Pant, Stripe
In this video, Mayank Pant from Stripe discusses the challenges and strategies of AI pricing, emphasizing the need for hybrid pricing models that adapt rapidly to changing features and customer needs, allowing companies to maintain competitive margins while fostering customer trust and engagement.