Suggest modification to this talk

Title

Description

Have you ever wanted to automate a tedious process? Like, relaying omelette photos straight from your phone to an omelette blog? This talk is for you! Learn how to leverage Ruby within serverless functions to handle everything from machine learning based object detection through event triggers, and static site generation!

As the system for automation is explored, so too are the challenges, or, brick walls, that were hit along the way. In this talk, you'll not only learn about technology, but also about mental models for solving problems and overcoming those inevitable brick walls.

Kevin Lesht
Kevin is a Senior Software Engineer working out of Chicago, Illinois, for Home Chef, a meal kit delivery company. He is the host of the Day as a Dev podcast, where he interviews developers from all walks of life, to offer students and those interested in tech, a practical look at life in the industry. Previously, he co-founded ZeroSleep Designs, a digital design agency, and worked with dozens of clients to bring their ideas online. He's passionate about getting others excited about tech!

Date

Summarized using AI?

If this talk's summary was generated by AI, please check this box. A "Summarized using AI" badge will be displayed in the summary tab to indicate that the summary was generated using AI.

Show "Summarized using AI" badge on summary page

Summary

Markdown supported

In the RubyConf 2020 talk titled "Automation Engineering with Serverless Compute," Kevin Lesht shares his journey towards automating the process of updating his omelette blog using serverless computing and Ruby. This hands-on session explores leveraging serverless functions to automate a tedious task that many encounter—managing content updates for a blog.

Key Points Discussed:  
- **Automation Goals**: Kevin commences with the challenge of posting omelette photos to his blog, explaining his desire for a seamless process where a photo taken automatically gets uploaded if it's of an omelet.  
- **Identifying Structures**: To achieve this, Kevin identifies necessary components in the automation pipeline, including capturing photos, parsing and detecting if they are actually omelets, and uploading them to his blog.  
- **Using Google Photos API**: He thinks of utilizing the Google Photos API to directly fetch photos from his mobile devices but highlights the complexity of implementing OAuth for authorization without a web interface.  
- **Embracing Serverless Technology**: Kevin discusses the potential benefits of serverless computing (like AWS Lambda) to abstract away infrastructure management, simplifying his focus on the code.  
- **Brick Walls**: As he navigates challenges (or "brick walls"), Kevin emphasizes persistence and the importance of problem-solving in software development. He cites inspiration from Randy Posh's insights on overcoming obstacles.  
- **Leveraging AWS Recognition**: To facilitate object detection in photos, Kevin explains his use of AWS Recognition, which initially involves using pre-trained models before realizing the necessity for a custom model to accurately detect omelets.  
- **Building a Custom Model**: He elaborates on the process of creating a reliable dataset with labeled images to successfully train a model for recognizing an omelette.  
- **Automation Strategy**: Through a series of serverless functions including parsing photos, processing them concurrently with Lambdas, and storing relevant timestamps in DynamoDB to keep track of submissions, Kevin elaborates the creation of an automated workflow that efficiently updates the blog.  
- **Final Thoughts**: In closing, he acknowledges that while the implemented solution might be complex, the real learning comes from embracing challenges and recognizing "brick walls" as opportunities for growth. Kevin hints at future projects, showing his enthusiasm for continuous learning in the tech space.

Ultimately, Kevin's presentation illustrates how innovative serverless technology can streamline tasks and the importance of resilience when solving technical problems.

Cancel