CloudSaver AI - Scan your AWS bill CSV for waste in seconds — free

by
CloudSaver AI scans your AWS Cost Explorer CSV for wasted spend in seconds. No API keys, no signup, no data leaving your browser. Detects zombie resources, over-provisioned EC2, idle load balancers, stale EBS snapshots, unused Elastic IPs, and S3 tier mismatches. Built for indie hackers and small SaaS teams paying $200-5000/month to AWS. Free scan now. Pro (auto-scans + alerts) is on the roadmap — leave your email on the site to get early access.

Add a comment

Replies

Best
Maker
📌
I spent years watching AWS bills grow while everyone assumed it was "just the cost of doing business." The truth? Most small teams waste 28-35% of their cloud spend on resources nobody is using — orphaned IPs, oversized instances, idle load balancers, stale snapshots. The existing tools (CloudHealth, Vantage, Infracost) are great, but they're built for enterprises with DevOps teams. I wanted something for indie hackers, solo founders, and small SaaS teams who pay $500-5000/month and don't have time to become FinOps experts. So I built CloudSaver AI: drop your AWS Cost Explorer CSV, get a waste report in 3 seconds. No API keys. No signup. No data leaves your browser. This is an MVP. The current engine is rule-based, but I plan to add AI reasoning once I validate demand. I'd love your feedback — what waste patterns am I missing?

Hey Product Hunt! 👋

I'm Malin, solo founder of CloudSaver AI. I built this after years of seeing the same waste patterns in every AWS bill — idle IPs, oversized instances, forgotten snapshots. The big FinOps tools are built for enterprises with dedicated DevOps teams. I wanted something any indie hacker or small team could use in 10 seconds.

A few things I'm proud of:
- Your billing CSV never leaves your browser. There's literally no backend. I can't see your data even if I wanted to.
- No IAM roles or access keys required. Just export your Cost Explorer CSV and drop it in.
- The whole thing is a single HTML file. No dependencies, no npm install, no Docker.

What I'd love feedback on:
1. Are there waste patterns I'm missing? (Currently detecting 6 types)
2. Would you trust a browser-only tool with your billing data?
3. Would you join a waitlist for monthly auto-scans + an optimization playbook?

Try the demo (auto-loads sample data, no upload needed):

Happy to answer any technical questions about how the detection engine works!

actually ran my last month's CSV through it and it caught an idle load balancer i totally forgot about, kind of embarrassed honestly. the fact that nothing leaves your browser is a really nice touch too

 Hülya, this made my day 🙏

An idle LB caught on the first scan — that's exactly why I built this. Those things silently cost $15-30/month and everyone forgets about them.

Quick question if you don't mind: how was the upload experience? Any confusing parts or things you wished were different? I'm still learning what works for real users vs. what I think works.

Also — would you be open to a quick follow-up in a week to see if the savings tip actually helped? Trying to validate whether people act on the results or just look at them once.

the fact that everything runs client-side without uploading your cost data is a really thoughtful call, especially for smaller teams who are wary of yet another SaaS tool asking for credentials. loving the focused scope too.

 Thanks Egeman! That was actually the #1 design decision — I kept seeing tools that want read-only IAM roles or full account access, and even though I understand why, it's a big ask for a 10-second check-in.

The trade-off is I can't do continuous monitoring (no backend = no cron jobs), but for a "quick health check" vibe I think the browser-only model works. Curious to hear your thoughts on where the line should be — at what point does a tool become useful enough to justify giving it API access?

Ran my last quarter's CSV through it and it flagged three EBS snapshots I'd totally forgotten about. Love that nothing leaves the browser, way less friction than wiring up API keys for a quick audit.

 Thanks Ertuğrul! Really appreciate you sharing the actual use case — 3 forgotten EBS snapshots is exactly the silent waste we built this for. Glad the no-API-key, browser-only approach worked for you (that was a deliberate call, not a limitation).

We're shaping a Pro direction: scheduled scans + alerts so you don't have to remember to run it each quarter. Genuinely curious — would monthly automated AWS waste reports be worth paying for to you? Early access or a custom audit → just email me at . Either way, thanks for actually trying it out.

Would be great if you could add multi-account support since a lot of small SaaS teams run separate AWS accounts per environment or client project. Even just letting us upload multiple CSVs at once and showing combined waste across them would make this way more useful than running one scan per account.

 Thanks Özlem! Multi-account is one of the most requested features on our roadmap. The plan: batch CSV upload → unified waste dashboard across accounts. It's absolutely doable on the browser-only model (no API keys needed).

If you'd like early access to the multi-account version when it ships, drop me a line at — I'll keep you posted.

would love to see an option to upload multiple months of cost explorer exports so it can flag resources that have been sitting idle for 30, 60, or 90+ days instead of just one snapshot

 Thanks Asel! Multi-month comparison is on our radar — being able to tag resources as idle-for-30/60/90 days would catch waste that a single snapshot misses. We're planning batch CSV upload support next. Want early access when it ships? Drop me a line at .

The CSV upload flow is solid, but adding a way to compare two scans side by side would be huge. Drop in last month's export alongside this month's and instantly see which zombie resources disappeared or which new costs popped up. Helps prove ROI and spot regressions in one glance.

 Sevim — spot-on about ROI proofing. A month-over-month diff view (what disappeared vs what's new) is exactly where we're heading after the single-scan MVP. That "regression spotting" use case is something we hadn't framed as clearly as you just did — appreciate that. Email me at if you want early access to the comparison version.

Finally something that just reads my CSV without asking for an IAM key, the zombie EC2 and stale snapshot detection caught two things I'd forgotten about. Pretty handy for a quick gut check before the bill lands.

 Buglem — love hearing it actually caught something real! Two forgotten resources is exactly what this was built for. If you end up running it monthly before bills drop, that's a use case we'd love to build around (think: scheduled reminders + trend tracking). if you want in on that.

A few honest notes from the maker:

Launch day's most valuable signal wasn't the upvotes — it was the feedback. Several people ran their own AWS Cost Explorer CSV through CloudSaver and found resources they'd forgotten about: idle EC2 instances, stale EBS snapshots, that kind of silent waste. Exactly the problem we built this for.

The most-requested next step, unprompted from multiple people: multi-account support and month-over-month comparison (flag idle-for-30/60/90 days, spot regressions). So that's what Pro is shaping into — same browser-only model, no API keys.

Want early access when it ships, or a custom AWS audit now? Email me directly: . No signup wall.

Thanks to everyone who actually tried it and gave real feedback.