Comment by skyzouwdev

I like the premise — I’ve lost count of how many times I ended up deep in Google’s search results or on page 4 of some company blog just to find a relevant architecture post.

A couple of thoughts from my own workflow: • I’d definitely add smaller but high-signal engineering blogs. Some of the most interesting write-ups I’ve read came from companies that aren’t FAANG-level but are operating at scale (think Segment, Plaid, Posthog, Linear). They often publish very focused “lessons learned” pieces that don’t get much SEO love. • AI summaries could be useful if they’re really accurate and highlight the “why” behind a technical decision, not just rephrase the intro paragraph. I’d probably use them to triage which posts to read in full. • For discovery, I usually rely on Twitter/X and the occasional Reddit thread. The problem is that those channels are noisy — so something that’s searchable and filterable like you’ve built is appealing.

One nit: for tagging, maybe consider a lightweight NLP approach (spaCy / transformers) that can detect concepts beyond exact matches. Even basic keyword clustering could improve relevance without needing a massive ML pipeline.

Nice work — bookmarking this for the next time I need to dig into how someone scaled their job queues without setting their hair on fire.