About Summa Reviews
What is Summa Reviews?
You know that feeling when you want to buy something and end up watching 47 YouTube reviews at 2x speed while your coffee gets cold? Yeah, I fixed that.
Summa Reviews aggregates real human opinions from YouTube reviewers and presents them in a clean, categorized Pros & Cons table. No AI hallucinations here — every single point comes from actual humans who actually used the product and had actual opinions about it. These reviewers put in great work testing products and sharing their honest experiences — I just made it easier to access all their insights in one place.
How it works (it's embarrassingly simple):
3 clicks to enlightenment: Find a product → Click any pro or con → Get teleported to the exact timestamp in the original review where that point was made. Want to know about "battery charging speed" for that tablet? Click. Done. No more scrubbing through 23 minutes of unboxing footage.
Quick Ranking: Just want a number? I got you. Each product gets an overall score based on the ratio of pros to cons mentioned across all reviews I analyzed (on average, that's about 18 videos per product). It's like Rotten Tomatoes, but for whether that air fryer actually makes crispy fries.
Smart Comparisons: Torn between three different air fryers? Compare them side-by-side and see how each one scores in "Cooking Performance," "Ease of Cleaning," and other categories. Real numbers from real reviewers, not vibes.
Got Suggestions?
I genuinely want to hear from you. Feature ideas, bug reports, complaints about my algorithm's questionable life choices — all welcome.
*Unlike Michael, I promise to actually read your suggestions.*
Keep This Project Running
Real talk: this is a passion project, but it's not free to run. Processing tens of thousands of videos through LLMs costs real money, and I've spent two years of evenings and weekends building this.
If Summa Reviews helps you make a purchase decision, consider using the Amazon affiliate links on the product pages. It doesn't cost you anything extra, but it helps me cover the costs and keep improving the service. Every little bit helps keep the servers humming and the algorithms learning.
Or if you're feeling generous:
No pressure though — I built this because I needed it, and I'm happy you're here either way.
The Person Behind This
Hi! I'm a neuroscientist and data scientist who apparently has a strange definition of "free time activities."
This whole thing started because I'm *that person* who can't buy anything without watching at least 10 reviews first. Smartphone? 15 reviews minimum. Gaming mouse? Gotta compare every sensor. Air fryer? Don't even get me started. Electric car? I basically wrote a dissertation before test-driving one.
The problem? I have a full-time job and only so many evenings to sacrifice to the YouTube algorithm. So I thought: *"What if I could extract all the useful bits from reviews automatically?"*
Two years of weekends and late nights later, here we are. Summa Reviews is my hobby project that got slightly out of hand (in the best way). It's built for people like me — people who want to make informed decisions but don't have unlimited time to consume content about every purchase.
I made this myself, which means it's crafted with love but might occasionally have a bug or two. If you spot something weird, there's a Report button on every product page. Please use it! I actually read those reports.
The Engine
Ah yes, the algorithm. My beautiful, overcomplicated child.
Here's the thing about YouTube reviews: not all of them are useful. Some are AI-generated garbage. Some are just someone reading the spec sheet. Some compare 17 products in one video (looking at you, "BEST PHONES 2024 ULTIMATE MEGA COMPARISON").
So the algorithm has to work *hard*. It goes through multiple cleaning passes to filter out the noise, remove gibberish, and actually understand what's a genuine pro or con versus just filler content. Every piece of data passes through the LLM multiple times — once to extract, once to clean, once to categorize, and sometimes a few more times because it got something hilariously wrong the first time.
The categorization is smart: Different products get different categories. Smartphones get Camera, Display, Performance, Battery, etc. Hair dryers get Drying Speed, Heat Settings, Hair Health, Styling Capabilities. Air fryers get Crispiness (yes, that's a technical category now). The algorithm figures out what matters for each product type.
The comparison video problem: Oh boy. Do you know how hard it is to process a video called "iPhone vs Samsung vs Pixel vs OnePlus ULTIMATE BATTLE" and extract info about *just* the iPhone? The algorithm has to make multiple extra passes to isolate which comments belong to which product. It was... a journey.
It was a bumpy ride with countless edge cases and "why is this happening" moments. But hopefully, I've brought you something actually useful. If there are still mistakes (and there will be some, because automation), I sincerely apologize. Hit that report button. I'll fix it.
What's Coming Next
Your voice matters! Here's what's on the roadmap:
- Vote for Products: Want me to analyze a specific product or category? Soon you'll be able to vote for what gets processed next. Democracy in action.
- Real Prices: Because knowing a laptop is great means nothing if you can't afford it. I'm working on adding actual pricing data so you can make decisions with your wallet in mind.
- More Categories: If it has reviews on YouTube, I can probably analyze it. Suggest categories you'd like to see!
*Thanks for using Summa Reviews. Now go make an informed purchase decision and actually enjoy your warm coffee for once.*
Questions? Reach me at info@summareviews.ai