BACK

AI Usage Framework and Personas
- AI usage involves: context switching, prompting, reviewing output, and fixing AI-generated results.
- Three personas introduced for effective AI use:
1. Efficient Edith: Use AI when the total time (switching, prompting, reviewing) is less than manual effort.
2. Frequent Felix: Invest time to build reusable AI workflows/tools for repeated problems, which pays off over time.
3. Quick Quinton: Use AI for small, quick tasks that save time without heavy context or editing.

Efficient Edith Use Cases
- Writing lint rules (example: converted a 4-hour task into 4 minutes by prompting AI to write code plus tests).
- Creating one-off scripts for tasks like generating failing file lists for linting; these scripts save large amounts of manual work.
- Diagramming: Convert text into system diagrams using Excalibur AI or Mermaid syntax for proposals and technical docs in minutes.

Frequent Felix Use Cases
- Repetitive data transformation: Using AI to generalize SQL query adjustments over many queries, reducing hours to minutes.
- Building internal tools: Example of a testing tool that generates accurate tests from code context, saving weeks of work and thousands of engineering hours.
- Creating UI tools for repeated tasks like Markdown to HTML conversion accessible to all engineers.
- Reusable prompts like writing coaches and documentation generators to improve text quality and speed.
- Managers benefiting from tailored AI feedback tools for their engineers to streamline review processes.

Quick Quinton Use Cases
- Fast, small help on shell commands: AI-assisted creation of complex shell commands and helper functions to automate common workflows.
- Debug statements generation: Adding comprehensive logs automatically saves debugging time.
- Library and ecosystem exploration: Using AI to find appropriate libraries and quickly learn unfamiliar technology stacks to expedite contributions.

Recap and Final Advice
- AI is transforming workflows with potential 2x to 100x productivity gains depending on task.
- Encourages viewing work through the lens: “Can AI get me most of the way there?”
- Know when to apply AI using personas for maximum value.
- Sharing knowledge and tools boosts team productivity.
- Jordan offers to discuss AI further and shares resources via LinkedIn and slides.

Actionable Items / Tasks
- Evaluate your tasks to identify which fit Efficient Edith, Frequent Felix, or Quick Quinton personas for AI application.
- Use AI to automate cumbersome coding tasks like lint rule writing, scripting, and diagram generation.
- Build or adopt reusable AI-powered internal tools for repeatable tasks.
- Create reusable AI prompts, e.g., for documentation and writing quality improvements.
- Use AI as a quick assistant for shell commands, debugging, and exploring unfamiliar codebases.
- For managers: consider developing or adopting AI tools that assist engineers in self-service feedback to improve review efficiency.
- Continuously assess your workflows for AI integration opportunities to achieve time savings and productivity increases.

Using AI to 10x Your Productivity as an Engineer

Share:

14:10 - 14:40, 28th of May (Wednesday) 2025 / DEV TRENDS STAGE

AI is changing the engineering industry faster than most can keep up. In this talk, Jordan teaches frameworks to think about when and why to use AI, along with many practical examples of how you can use AI to be more productive based on his experience as a Senior Engineer at Pinterest. The practical strategies will be a mix of coding and non-coding-related tasks common to engineers every day. If you're an engineering leader, you'll still find the frameworks and strategies useful, plus you can pass on your learnings to your team to make them more productive!

LEVEL:
Basic Advanced Expert
TRACK:
AI/ML Dev Software
TOPICS:
AI FutureTrends SoftwareEngineering WebDev

Jordan Cutler

Pinterest