ai that works.

Real AI systems, built live every Tuesday. Agents, evals, RAG, and what actually holds up in production.

with Dex Horthy (HumanLayer) & Vaibhav Gupta (Boundary)

YouTube GitHub

78episodes

1:15:18
Why Performance Engineering Breaks AI Coding Agents
tuesday-seriesSep 18, 2026
55:02
Skip the Sandbox. Give Every Engineer a Mac Mini.
context-engineeringSep 11, 2026
1:10:47
Code Mode: Why AI Agents Are Ditching Tool Calls for Code
production-aiSep 4, 2026
1:10:45
How to Build a Software Factory for AI Coding Agents
Aug 28, 2026
18:19
The 1995 Programming Language That Already Invented AI Agents
Aug 26, 2026
58:38
The Data Structure Problem AI Coding Agents Can't See
Aug 21, 2026
49:31
AI Engineers Can't Agree on One Workflow. Here's Why.
Aug 14, 2026
52:41
Can AI Coding Agents Actually Build Maintainable Software?
Aug 7, 2026
52:08
How to Test New AI Models Before They Break Production
Latency & CostJul 31, 2026
55:22
Building a Shared Memory System for AI Coding Agents
Coding AgentsJul 24, 2026
46:10
Why Passing Benchmarks Doesn't Mean Your AI Wrote Good Code
Coding AgentsJul 17, 2026
52:23
Stop Reading Code. Start Understanding Systems.
Coding AgentsJul 10, 2026
51:56
The Self-Healing Agent Loop That Fixes Its Own Language
Agent ArchitectureJun 26, 2026
51:10
Why Your AI Coding Agent Keeps Writing Bad Code (It's Not the Prompt)
Coding AgentsJun 19, 2026
42:28
Testing Claude Fable 5: Why New AI Models Rarely Change Everything
Coding AgentsJun 13, 2026
1:02:08
Why Performance Engineering Still Requires Human Judgment | No Vibes Allowed
Coding AgentsJun 12, 2026
38:03
How AI Agents Can Safely Ship Code to Production
Coding AgentsJun 5, 2026
52:52
How to Build AI Agents That Work in Any Language
Prompt EngineeringJun 5, 2026
1:00:31
Why Bash Might Be Holding AI Agents Back | Rhys Sullivan
Agent ArchitectureJun 1, 2026
52:12
Can You Outsmart the Model Makers?
Agent ArchitectureMay 11, 2026
4:31
Save 20-50% on agent tool call cost and latency
Tools & MCPMay 5, 2026
47:22
Can an AI Out-Plan a Senior Engineer?
Coding AgentsMay 4, 2026
57:05
Context Engineering vs Harness Engineering vs Software Engineering
Context EngineeringApr 27, 2026
52:59
Building a Practical AI Assembly Line
AI Product UXApr 20, 2026
59:51
Streaming Systems Masterclass
Coding AgentsApr 13, 2026
2:02:26
Build Faster by Coding Slower
Coding AgentsApr 6, 2026
9:00
Planning is dead: Recursive order-n planning
Coding AgentsApr 4, 2026
58:52
The Right Way to Give Your AI New Abilities
Tools & MCPMar 30, 2026
58:00
Prompt-Hackers are coming for your data
Security & GuardrailsMar 23, 2026
1:00:05
Agents, Subagents, Skills and Commands
Context EngineeringMar 16, 2026
1:03:09
How to Scrub Sensitive Data Before it Reaches Your LLM
Security & GuardrailsMar 9, 2026
51:16
The No-Rework Workflow for AI Coding Assistants
Coding AgentsMar 2, 2026
59:47
How to Automate Complex Workflows with Claude: 🦄 #45
AI WorkflowsFeb 23, 2026
52:51
Agentic Backpressure Deep Dive: 🦄 #44
Agent ArchitectureFeb 16, 2026
50:14
Prompting Is Becoming a Product Surface: 🦄 #43
AI Product UXFeb 9, 2026
1:42:47
Claude Code Maxing - live coding: 🦄 #42
Coding AgentsFeb 2, 2026
58:32
Email is All You Need: 🦄 #41
Agent ArchitectureJan 26, 2026
1:14:03
Applying 12-Factor Principles to Coding Agent SDKs:🦄 #40
Agent ArchitectureJan 19, 2026
1:04:01
Understanding Latency:🦄 #39
Latency & CostJan 12, 2026
1:08:10
Founding Boundary: 🦄 #38
Founder StoriesJan 5, 2026
53:46
Founding HumanLayer: 🦄 #37
Founder StoriesDec 29, 2025
1:00:29
Building a Prompt Optimizer:🦄 #36
Prompt EngineeringDec 22, 2025
54:54
Git Worktrees + Agents:🦄 #35
Coding AgentsDec 15, 2025
1:01:19
Multimodal Evals:🦄 #34
Evals & TestingDec 8, 2025
2:22:47
No Vibes Allowed: 🦄 #33
Live CodingDec 2, 2025
52:03
Building Animation Pipelines: 🦄 #32
Coding AgentsNov 25, 2025
58:02
Date, Times, and LLMs: 🦄 ai that works #31
Prompt EngineeringNov 12, 2025
58:06
Event Driven Agent Loops: 🦄 #30
Agent ArchitectureNov 5, 2025
1:18:06
🦄 ai that works: Ralph Wiggum under the hood: Coding Agent Power Tools
Context EngineeringOct 29, 2025
1:18:53
Agentic Rag: Building a coding agent (no frameworks) 🦄 ep #28
Context EngineeringOct 22, 2025
2:07:42
No vibes allowed (AKA live coding with Claude and Code Layer): 🦄 #27
Live CodingOct 16, 2025
59:10
Anthropic Post Mortem: 🦄 #26
Latency & CostOct 16, 2025
1:00:31
🦄 Dynamic Schemas: #25
AI Product UXOct 4, 2025
1:09:02
🦄 Evals for large scale classification: #24
Evals & TestingOct 4, 2025
53:04
🦄 Bash vs MCP for Coding Agents: ep #23
Tools & MCPOct 4, 2025
56:12
🦄 ai that works: Generative UIs and Structured Streaming
AI Product UXSep 10, 2025
54:46
Voice Agents and Supervisor Threading - 🦄 #21
AI Product UXSep 6, 2025
1:01:22
Claude for non-code tasks: 🦄 #20
AI WorkflowsAug 28, 2025
1:01:07
Interruptible agents: 🦄 #19
Agent ArchitectureAug 20, 2025
1:14:45
Context Engineering lessons from Manus - 🦄 #18
Latency & CostAug 16, 2025
1:11:56
Advanced context engineering for coding agents: 🦄#17
Coding AgentsAug 6, 2025
1:20:02
Evals: How to compare models 🦄#16
Evals & TestingAug 2, 2025
1:01:44
🦄 ai that works: PDFs, Multimodality, Vision Models: Part 1
Document ExtractionJul 27, 2025
11:53
🦄 ai that works: PDFs, Multimodality, Vision Models
Document ExtractionJul 23, 2025
1:26:29
Implementing Decaying-Resolution Memory: 🦄 #14
Agent ArchitectureJul 16, 2025
1:31:20
🦄 ai that works: Context Engineering and memory deep dive #13
Context EngineeringJul 9, 2025
1:15:45
Getting Tone Just right with LLMs: 🦄 #12
Prompt EngineeringJul 4, 2025
1:29:27
Using LLMs to go from 60+ min Youtube video to email / X posts: 🦄 #11
AI WorkflowsJun 27, 2025
1:07:50
Entity extraction from LLMs - extracting, deduping, enriching -🦄 Ep #10
Document ExtractionJun 20, 2025
1:23:48
Cracking the prompting interview: 🦄 Ep #9
Prompt EngineeringJun 13, 2025
1:12:40
Human as Tools: 🦄 Ep #8
Tools & MCPJun 10, 2025
1:14:44
Using MCP server with 10000+ tools: 🦄 Ep #7
Context EngineeringMay 28, 2025
1:18:30
LLMs to analyze Enron Emails: 🦄 Ep #6
Document ExtractionMay 21, 2025
1:27:41
🦄: Evals, Evals Evals - Episode #5
Evals & TestingMay 13, 2025
1:27:54
🦄 Building a 12 Factor Agent: ep #4
Agent ArchitectureApr 22, 2025
1:01:21
🦄 building a coding bot: ep#3
Coding AgentsApr 15, 2025
1:16:42
🦄 reasoning models vs reasoning prompts: ep#2
Prompt EngineeringApr 7, 2025
1:12:27
🦄 large scale classification: ep#1
Document ExtractionApr 2, 2025

ai that works.

Why Performance Engineering Breaks AI Coding Agents

Skip the Sandbox. Give Every Engineer a Mac Mini.

Code Mode: Why AI Agents Are Ditching Tool Calls for Code

How to Build a Software Factory for AI Coding Agents

The 1995 Programming Language That Already Invented AI Agents

The Data Structure Problem AI Coding Agents Can't See

AI Engineers Can't Agree on One Workflow. Here's Why.

Can AI Coding Agents Actually Build Maintainable Software?

How to Test New AI Models Before They Break Production

Building a Shared Memory System for AI Coding Agents

Why Passing Benchmarks Doesn't Mean Your AI Wrote Good Code

Stop Reading Code. Start Understanding Systems.

The Self-Healing Agent Loop That Fixes Its Own Language

Why Your AI Coding Agent Keeps Writing Bad Code (It's Not the Prompt)

Testing Claude Fable 5: Why New AI Models Rarely Change Everything

Why Performance Engineering Still Requires Human Judgment | No Vibes Allowed

How AI Agents Can Safely Ship Code to Production

How to Build AI Agents That Work in Any Language

Why Bash Might Be Holding AI Agents Back | Rhys Sullivan

Can You Outsmart the Model Makers?

Save 20-50% on agent tool call cost and latency

Can an AI Out-Plan a Senior Engineer?

Context Engineering vs Harness Engineering vs Software Engineering

Building a Practical AI Assembly Line

Streaming Systems Masterclass

Build Faster by Coding Slower

Planning is dead: Recursive order-n planning

The Right Way to Give Your AI New Abilities

Prompt-Hackers are coming for your data

Agents, Subagents, Skills and Commands

How to Scrub Sensitive Data Before it Reaches Your LLM

The No-Rework Workflow for AI Coding Assistants

How to Automate Complex Workflows with Claude: 🦄 #45

Agentic Backpressure Deep Dive: 🦄 #44

Prompting Is Becoming a Product Surface: 🦄 #43

Claude Code Maxing - live coding: 🦄 #42

Email is All You Need: 🦄 #41

Applying 12-Factor Principles to Coding Agent SDKs:🦄 #40

Understanding Latency:🦄 #39

Founding Boundary: 🦄 #38

Founding HumanLayer: 🦄 #37

Building a Prompt Optimizer:🦄 #36

Git Worktrees + Agents:🦄 #35

Multimodal Evals:🦄 #34

No Vibes Allowed: 🦄 #33

Building Animation Pipelines: 🦄 #32

Date, Times, and LLMs: 🦄 ai that works #31

Event Driven Agent Loops: 🦄 #30

🦄 ai that works: Ralph Wiggum under the hood: Coding Agent Power Tools

Agentic Rag: Building a coding agent (no frameworks) 🦄 ep #28

No vibes allowed (AKA live coding with Claude and Code Layer): 🦄 #27

Anthropic Post Mortem: 🦄 #26

🦄 Dynamic Schemas: #25

🦄 Evals for large scale classification: #24

🦄 Bash vs MCP for Coding Agents: ep #23

🦄 ai that works: Generative UIs and Structured Streaming

Voice Agents and Supervisor Threading - 🦄 #21

Claude for non-code tasks: 🦄 #20

Interruptible agents: 🦄 #19

Context Engineering lessons from Manus - 🦄 #18

Advanced context engineering for coding agents: 🦄#17

Evals: How to compare models 🦄#16

🦄 ai that works: PDFs, Multimodality, Vision Models: Part 1

🦄 ai that works: PDFs, Multimodality, Vision Models

Implementing Decaying-Resolution Memory: 🦄 #14

🦄 ai that works: Context Engineering and memory deep dive #13

Getting Tone Just right with LLMs: 🦄 #12

Using LLMs to go from 60+ min Youtube video to email / X posts: 🦄 #11

Entity extraction from LLMs - extracting, deduping, enriching -🦄 Ep #10

Cracking the prompting interview: 🦄 Ep #9

Human as Tools: 🦄 Ep #8

Using MCP server with 10000+ tools: 🦄 Ep #7

LLMs to analyze Enron Emails: 🦄 Ep #6

🦄: Evals, Evals Evals - Episode #5

🦄 Building a 12 Factor Agent: ep #4

🦄 building a coding bot: ep#3

🦄 reasoning models vs reasoning prompts: ep#2

🦄 large scale classification: ep#1