Google Gemini Writing Challenge

Source: Dev.to

What I Built

• Where Gemini fit in
  • Used Gemini’s multimodal capabilities to let users upload screenshots of notes, diagrams, or code snippets.
  • Gemini generated structured study plans, summaries, and practice questions based on the uploaded content.
  • Deployed the backend on Cloud Run, which made scaling effortless and kept the API layer clean and lightweight.
  • (If this were a real DEV post, embed the Cloud Run link here.)

What Surprised Me

• Gemini handled messy input better than expected. Screenshots with scribbles, half‑written code, or low‑contrast text still produced surprisingly accurate summaries.
• Prompt engineering mattered less than anticipated; concise, intention‑focused instructions worked best.
• Latency remained consistently low, even when processing images, making the assistant feel responsive enough for real study sessions.

What I Learned

Technical skills

• Became more comfortable with Cloud Run’s deployment flow and service revisions.
• Designed a lightweight API wrapper around Gemini that handles multimodal requests without tangled conditionals.
• Adopted better patterns for caching and rate‑limiting when working with AI APIs.

Soft skills

• Learned to scope aggressively; trimming the original oversized idea improved the final product.
• Improved at user testing—observing others use the tool provides humbling, valuable feedback.

Unexpected lessons

• Users often don’t know what they want from an AI tool until they see it.
• The most‑loved feature was a last‑minute addition: “Explain this like I’m tired,” which provides a super‑simple summary.

My Honest Thoughts on Google Gemini

• Multimodal support is genuinely impressive.
• The API is straightforward and predictable.
• The model feels more “context‑aware” than others I’ve used, especially with mixed text + images.

Where I Hit Friction

• Some responses were overly verbose, even when concise output was requested.
• Image‑understanding quality occasionally dipped with handwritten notes.
• Error messages could be clearer; it was sometimes unclear whether an issue stemmed from the request or the model.

None of these were deal‑breakers, but they slowed me down enough to notice.

What’s Next

• Expand the assistant into a more complete learning companion that tracks progress, adapts difficulty, and possibly integrates spaced repetition.
• Leverage Gemini’s strong foundation to continue exploring new possibilities.