One of the biggest challenges was getting our backend to work smoothly on AWS Lambda. Local testing was fine, but once deployed we hit issues like timeouts, cold starts, missing permissions, and even ES module errors because Lambda handles imports differently. Debugging these problems took time because the logs weren’t always clear. We fixed it by breaking the code into smaller pieces, rewriting the handler to be fully ESM-compatible, and carefully checking IAM permissions and payload sizes.

Another hurdle was fetching external data. APIs like Google News RSS and GNews often returned inconsistent formats or blocked the request entirely. Some articles didn’t have images, some had broken links, and a few endpoints had rate limits. We solved this by writing a fallback parser, adding validation for missing fields, and building a small caching layer so the frontend didn’t depend on live API calls every time.

Finally, integrating the frontend with the Lambda endpoints required a bit of wrestling with CORS. Our first few attempts caused blocked requests. After checking Amplify hosting settings and adding proper CORS headers in the Lambda response, everything worked smoothly.

Each of these issues slowed us down, but solving them gave us a much deeper understanding of AWS and improved the reliability of the project.