Your Support Team Should Ship Code – Lisa Orr, Zapier

AI Engineer · Beginner ·🛠️ AI Tools & Apps ·5mo ago
Zapier maintains 8000+ integrations that break as APIs change. We had thousands of backlog support tickets with dozens more arriving weekly. To keep up with the traffic, we started building AI tools to help ship integration fixes faster. We began by shadowing engineers fixing tickets and building tools we believed would expedite the fix process. Our first effort, an API playground hosting AI tools like diagnosis and test generation, failed to get engineering traffic because it pulled builders out of their workflows. We pivoted to MCP tools that engineers could use directly in their IDEs. MCP tools gained traction, but our most valuable tool, Diagnosis, took too long to run. Engineers wouldn't wait for it, revealing we needed an asynchronous approach. We built Scout Agent to string our tools together, autonomously reading support tickets, gathering context, generating fixes with tests, and submitting merge requests ready for review. This agent approach has gained traction with our support team handling high ticket volumes. An MR ready for review means they can validate and ship a fix quickly before needing to jump on the next incoming ticket. Throughout this process we've learned that the real challenge is everything surrounding code generation. Before writing code, Scout Agent needs both the right context and to show its work so engineers trust its recommendations. After generation, engineers need to quickly validate and correct the proposed fix, otherwise MRs sit unreviewed and abandoned. Embedding Scout Agent directly in GitLab solved this. Teams can iterate on proposed solutions without context switching. To track improvement, we measure three distinct failure modes: categorization accuracy (should Scout attempt this ticket?), fixability assessment (does this need a code fix?), and solution quality (does the generated code actually work?). Each reveals different improvement opportunities. Today, Scout drives 40% of support's integration fixes, with expansion to e
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Playlist

Uploads from AI Engineer · AI Engineer · 0 of 60

← Previous Next →
1 AI Engineer Summit 2023 — DAY 1 Livestream
AI Engineer Summit 2023 — DAY 1 Livestream
AI Engineer
2 AI Engineer Summit 2023 — DAY 2 Livestream
AI Engineer Summit 2023 — DAY 2 Livestream
AI Engineer
3 Principles for Prompt Engineering - Karina Nguyen (Claude Instant @ Anthropic)
Principles for Prompt Engineering - Karina Nguyen (Claude Instant @ Anthropic)
AI Engineer
4 Announcing the AI Engineer Network: Benjamin Dunphy
Announcing the AI Engineer Network: Benjamin Dunphy
AI Engineer
5 The 1,000x AI Engineer: Swyx
The 1,000x AI Engineer: Swyx
AI Engineer
6 Building AI For All: Amjad Masad & Michele Catasta
Building AI For All: Amjad Masad & Michele Catasta
AI Engineer
7 The Age of the Agent: Flo Crivello
The Age of the Agent: Flo Crivello
AI Engineer
8 See, Hear, Speak, Draw: Logan Kilpatrick & Simón Fishman
See, Hear, Speak, Draw: Logan Kilpatrick & Simón Fishman
AI Engineer
9 Building Context-Aware Reasoning Applications with LangChain and LangSmith: Harrison Chase
Building Context-Aware Reasoning Applications with LangChain and LangSmith: Harrison Chase
AI Engineer
10 Pydantic is all you need: Jason Liu
Pydantic is all you need: Jason Liu
AI Engineer
11 Building Blocks for LLM Systems & Products: Eugene Yan
Building Blocks for LLM Systems & Products: Eugene Yan
AI Engineer
12 The Intelligent Interface: Sam Whitmore & Jason Yuan of New Computer
The Intelligent Interface: Sam Whitmore & Jason Yuan of New Computer
AI Engineer
13 Climbing the Ladder of Abstraction: Amelia Wattenberger
Climbing the Ladder of Abstraction: Amelia Wattenberger
AI Engineer
14 Supabase Vector: The Postgres Vector database: Paul Copplestone
Supabase Vector: The Postgres Vector database: Paul Copplestone
AI Engineer
15 [Workshop] AI Engineering 101
[Workshop] AI Engineering 101
AI Engineer
16 The Hidden Life of Embeddings: Linus Lee
The Hidden Life of Embeddings: Linus Lee
AI Engineer
17 [Workshop] AI Engineering 201: Inference
[Workshop] AI Engineering 201: Inference
AI Engineer
18 The AI Pivot: With Chris White of Prefect & Bryan Bischof of Hex
The AI Pivot: With Chris White of Prefect & Bryan Bischof of Hex
AI Engineer
19 The AI Evolution: Mario Rodriguez, GitHub
The AI Evolution: Mario Rodriguez, GitHub
AI Engineer
20 Move Fast Break Nothing: Dedy Kredo
Move Fast Break Nothing: Dedy Kredo
AI Engineer
21 AI Engineering 201: The Rest of the Owl
AI Engineering 201: The Rest of the Owl
AI Engineer
22 Building Reactive AI Apps: Matt Welsh
Building Reactive AI Apps: Matt Welsh
AI Engineer
23 Pragmatic AI with TypeChat: Daniel Rosenwasser
Pragmatic AI with TypeChat: Daniel Rosenwasser
AI Engineer
24 Domain adaptation and fine-tuning for domain-specific LLMs: Abi Aryan
Domain adaptation and fine-tuning for domain-specific LLMs: Abi Aryan
AI Engineer
25 Retrieval Augmented Generation in the Wild: Anton Troynikov
Retrieval Augmented Generation in the Wild: Anton Troynikov
AI Engineer
26 Building Production-Ready RAG Applications: Jerry Liu
Building Production-Ready RAG Applications: Jerry Liu
AI Engineer
27 120k players in a week: Lessons from the first viral CLIP app: Joseph Nelson
120k players in a week: Lessons from the first viral CLIP app: Joseph Nelson
AI Engineer
28 The Weekend AI Engineer: Hassan El Mghari
The Weekend AI Engineer: Hassan El Mghari
AI Engineer
29 Harnessing the Power of LLMs Locally: Mithun Hunsur
Harnessing the Power of LLMs Locally: Mithun Hunsur
AI Engineer
30 Trust, but Verify: Shreya Rajpal
Trust, but Verify: Shreya Rajpal
AI Engineer
31 Open Questions for AI Engineering: Simon Willison
Open Questions for AI Engineering: Simon Willison
AI Engineer
32 Storyteller: Building Multi-modal Apps with TS & ModelFusion - Lars Grammel, PhD
Storyteller: Building Multi-modal Apps with TS & ModelFusion - Lars Grammel, PhD
AI Engineer
33 GPT Web App Generator - 10,000 apps created in a month: Matija Sosic
GPT Web App Generator - 10,000 apps created in a month: Matija Sosic
AI Engineer
34 Using AI to Build an Infinite Game: Jeff Schomay
Using AI to Build an Infinite Game: Jeff Schomay
AI Engineer
35 How to Become an AI Engineer from a Fullstack Background - Reid Mayo
How to Become an AI Engineer from a Fullstack Background - Reid Mayo
AI Engineer
36 The Code AI Maturity Model and What It Means For You: Ado Kukic
The Code AI Maturity Model and What It Means For You: Ado Kukic
AI Engineer
37 AI Engineer World’s Fair 2024 - Keynotes & Multimodality track
AI Engineer World’s Fair 2024 - Keynotes & Multimodality track
AI Engineer
38 From Text to Vision to Voice Exploring Multimodality with Open AI: Romain Huet
From Text to Vision to Voice Exploring Multimodality with Open AI: Romain Huet
AI Engineer
39 The Making of Devin by Cognition AI: Scott Wu
The Making of Devin by Cognition AI: Scott Wu
AI Engineer
40 The Future of Knowledge Assistants: Jerry Liu
The Future of Knowledge Assistants: Jerry Liu
AI Engineer
41 Llamafile: bringing AI to the masses with fast CPU inference: Stephen Hood and Justine Tunney
Llamafile: bringing AI to the masses with fast CPU inference: Stephen Hood and Justine Tunney
AI Engineer
42 Open Challenges for AI Engineering: Simon Willison
Open Challenges for AI Engineering: Simon Willison
AI Engineer
43 Lessons From A Year Building With LLMs
Lessons From A Year Building With LLMs
AI Engineer
44 From Software Developer to AI Engineer: Antje Barth
From Software Developer to AI Engineer: Antje Barth
AI Engineer
45 Unlocking Developer Productivity across CPU and GPU with MAX: Chris Lattner
Unlocking Developer Productivity across CPU and GPU with MAX: Chris Lattner
AI Engineer
46 Copilots Everywhere: Thomas Dohmke and Eugene Yan
Copilots Everywhere: Thomas Dohmke and Eugene Yan
AI Engineer
47 Fixing bugs in Gemma, Llama, & Phi 3: Daniel Han
Fixing bugs in Gemma, Llama, & Phi 3: Daniel Han
AI Engineer
48 Low Level Technicals of LLMs: Daniel Han
Low Level Technicals of LLMs: Daniel Han
AI Engineer
49 Emergence Launch: AI Agents and the future enterprise: Dr. Satya Nitta
Emergence Launch: AI Agents and the future enterprise: Dr. Satya Nitta
AI Engineer
50 How Codeium Breaks Through the Ceiling for Retrieval: Kevin Hou
How Codeium Breaks Through the Ceiling for Retrieval: Kevin Hou
AI Engineer
51 What's new from Anthropic and what's next: Alex Albert
What's new from Anthropic and what's next: Alex Albert
AI Engineer
52 Using agents to build an agent company: Joao Moura
Using agents to build an agent company: Joao Moura
AI Engineer
53 Decoding the Decoder LLM without de code: Ishan Anand
Decoding the Decoder LLM without de code: Ishan Anand
AI Engineer
54 Running AI Application in Minutes w/ AI Templates: Gabriela de Queiroz, Pamela Fox, Harald Kirschner
Running AI Application in Minutes w/ AI Templates: Gabriela de Queiroz, Pamela Fox, Harald Kirschner
AI Engineer
55 Building with Anthropic Claude: Prompt Workshop with Zack Witten
Building with Anthropic Claude: Prompt Workshop with Zack Witten
AI Engineer
56 Building Reliable Agentic Systems: Eno Reyes
Building Reliable Agentic Systems: Eno Reyes
AI Engineer
57 10x Development: LLMs For the working Programmer - Manuel Odendahl
10x Development: LLMs For the working Programmer - Manuel Odendahl
AI Engineer
58 Disrupting the $15 Trillion Construction Industry with Autonomous Agents: Dr. Sarah Buchner
Disrupting the $15 Trillion Construction Industry with Autonomous Agents: Dr. Sarah Buchner
AI Engineer
59 Hypermode Launch: Kevin Van Gundy
Hypermode Launch: Kevin Van Gundy
AI Engineer
60 Git push get an AI API: Ryan Fox-Tyler
Git push get an AI API: Ryan Fox-Tyler
AI Engineer

Related AI Lessons

Why You Accidentally Built a 5-App AI Stack
Learn how to avoid accidentally building a complex AI stack and simplify your workflow with focused tool integration
Dev.to · ForgeWorkflows
Doorbell Cameras Are Becoming Real-Time Awareness Systems
Doorbell cameras are evolving into real-time awareness systems with faster context and smarter alerts, enhancing home security
Medium · AI
I Built a Fully Automated YouTube Channel That Uploads Every Day - Without Touching It
Learn how to build a fully automated YouTube channel that uploads daily without manual intervention, leveraging AI and automation tools
Dev.to · Shuvo
I Didn’t Believe Free AI Tools Could Replace Paid Software… Until This Happened
Discover how free AI tools can replace paid software for content creation, saving time and increasing productivity
Medium · ChatGPT
Up next
China's New Qwen 3.7 is INSANE!
Julian Goldie SEO
Watch →