Migrating image processing from Vercel to AWS Lambda

What I worked on today

• Realized that my worker consumed half of my monthly Vercel usage overnight, making it unfit for production.
• Spent the entire day migrating the job to AWS Lambda, using S3 and BunnyCDN for cost-effective image storage and delivery.
• The migration is fully completed and working, making it a necessary step for long-term sustainability.

Lessons learned

• Vercel workers are not always cost-efficient for large-scale background jobs.
• Moving to AWS Lambda and BunnyCDN should lower costs and improve scalability.
• Some tasks feel like distractions, but they’re necessary for the bigger picture.

Challenges faced

• The migration took all day, leaving no time for feature work.
• Frustration over spending the entire day on infrastructure instead of feature development.
• Once again, I didn’t spend time on social media to promote Spyrate—it also feels like a waste of time at this stage.

What’s next

• Fix AI analysis, so I can generate "mock/real" data for the monitor activity detail.
• Once that’s done, focus on designing the monitor activity detail UI.