Amazon Real Time API

like this date last year (October 2024), I got obsessed about building an App to find Price errors on Amazon(like a product that costs normally $999 was mistakenly priced at $9,99). Since I have never wrote a single line of code in my life, I started doing researches, how can I achieve this? I learned that to achieve my goal to find such Price direpencies, I would need to continiously monitoring prices and offers of huge numbers of Amazon products. I have setup a goal and said if I monitored products from a 2 to 3 categories initially would be ok, but that meant around monitoring 10k products every 10 mn (because Price errors happens and quickly gets corrected ). then I spent like 2 months (October- December 2024) doing researches and using AI extensively (at that time ChatGPT was a trend). during these 2 months I have tested 10s of solutions like cloud scraping, third party scraping solutions, Proxy services and web blockers from Brightdata etc.. I was always like "no this is not what I am looking for" without even know what I am exactly looking for lol. and for my goal building the app will cost me a lot of mony to achieve millions of reqs/Month. (after calculations I made, the cheapes one of them will cost me at least $700-$1000 per 1 million request! )

The beginning of January 2025 I was almost hopeless, thenn wanted to give it another shot, I have signed up to a third party API in rapidapi, the guy and his Team are well known there as openwebninja, they have tier pricing at max tier 200k request for $150, I subscribed to it the using instructions from ChatGPT, I managed to create a workflow to get the Price errors and deals I was hunting. this gave me some hope back, but I noticed 2 issues quickly: the first was that they were caching responses (which means some data was not accurate and not in real time like they claimed) and the second was, I needed more quota than 200k/month. by the first week I have consumed all of the 200k already, I contacted them for more quota, but I didnt liked the suggested prices. Here I canceled my sub and wanted to build my own scraper again using AI+prompts. I followed the path of Browser automation using selenium/playright and all of that stuff, so the AI started genrating codes and I execute, till I was almost hopeless again because I faced the dilemma of needing proxies (which are incredibly high priced, for 1m Request I woud need to spend over a $1000) and never mention bot detection! + never like the idea of these browser automation stuffs cuz they were slow and made me to dependent to many uncessary things.

from March - June 2025; I stopped everything, a new member arrived to our family, my first son was born, I threw everything away but kept thinking about this.

from June - August: I got my passion again, this time even stronger, I forgot completey about my first driven cause (which is the price error app), and focused on how to make an Amazon scraper/extractor for large scale uses without depending on Browser automation and Proxies and that evade any kind of bot Protection implemented by the Amazon! this looked like impossible to happen, but I dived deep into that and the evolve of the AI models helped a lot. But even the newest AI models answered consistenly by the unlikeliness for such a system to be created. I insisted and used several AI models deepseek, grock, perplexity, claude,...etc to do my searches, and each time taking a response from this model and pasting it to the other one to investigate it and validate it. I have learned a lot of many techniques to evade bot detection systems and to create the Architecture of the system with under $150 per 1m call!

I kept saving and testing every idea I think it will work, till I crafted a plan of a workflow I learned from AIs, then bought a cheap vps, signed up to manus AI, prompt it to login to it and start coding according to the plan,it coded and debugged, not from a single prompt, but day in and day out I kept monitoring it and adjusting the plan while doing deep searches on problems we faced during the developement.

Somehow in late August it worked! I was like no way this can't be , the crafted system made over 4K successful requests without getting blocked in 4mins! for any technical expert would think ; from a single vps IP and 4K successful req is really hard to achieve! I even started battle testinng it with higher RPS, then I was able to achieve >50 RPS and all were successful! then I knew this is it! I have bypassed Amazon's bot detection and achieved what another scraping solution would take technical exxpertise and hundreds of dollars to achieve from a single vps, using AI and without me personnaly know how to code!

Since then and after I made the core files for networking, I started planing building the endpoints and the extractors. This one also was not easy, as it was also a huge learning curve to me about Parsing and extracting the data. I have learned it through the AI models during developement, my single point of strength is to never gaving up and pushing the AI to its limits! it can achieve it, it did !

By the 17th of semptember I have already designed the Aamzon API system with all its extractors/endpoints, deployed it to the cloud and finally to RapidAPI. I kept and still drafting it and perfecting some endpoints, Now I can say I literally have the most affordable and the cheapest Amazon API system in the market with max tier price at $150 per 400k request, and even with better data quality and dev friendly design.

Lessons learned:

• AI is a truly technoligical leap in our Age

• with AI you can code and debug complex APP heck without any prior coding knowledge.

• Never give up

  Guys I have made generous Free tier, feel free to test it out and give me your feeback please; https://rapidapi.com/keystonedata-keystonedata-default/api/amazon-real-time-api

  like I said, I am still perfecting the data quality and some filters for some endpoints, that , I will be happy to share to share with me your testing feedbacks and or if anything is not clear or if any one want more quota contact me at discord: https://discord.gg/GgNzySKy