Car price range recommendation tools development -- 2

I am developing a car marketplace where sellers register their vehicles to sell and we provide AI / Scientific/ Some Statistical Models by which can suggest/recommend an ideal price range $X-$Y way. It will be shown like our AI suggested price. eBay and many sites have this recommendation engine/ tool. For Example: Your Car: 2019 Toyota Camry LX, 20000 miles (User will pass us and we will have VIN# too). Our Recommended Sales Price: $18,500.00-$20,000.00 Every listing has VIN# so we know make, model, trim, year and we also have ZIP code of a car and seller’s email address along with other details. AI/ML/ BI or Data Driven Decision Making is not my technical strengths but from my common sense and past education/ experience I can say that this kind of system will have three components 1) Data Sources – We must have data to suggest or learn from – In my mind, we need to have our own data where we know about a car, sell price in our own inventory and other things. We might need to scrap. We need to connect to third party API providers in this space. We may need to buy data from others (DMV in USA – do they sell data for all sales price by VIN#)? [login to view URL], NADA, Edmunds, DMV, Inventory on hand, Competition/ Scrapper Data, Our own historic data (like eBay uses past data to suggest price for your listing) - all can chime in in my mind. 2) Intelligence – We cannot have just any kind of data. We need to understand what data will play a role in decision making/ intelligence. This will also drive the requirements if we need to buy data, partner with some companies (API providers), use our own data or collect more and/or scrap data from relevant sites. You need to show me a proof in your proposal that you can think some dependencies/ factors which should be fed into AI/Decision Making. In other words - How a car sell price recommendation should be changing? Like Competition Data is one source. What other factors should influence the range? I want to see how you think here. Most important in your proposal. 3) Model – Algorithm – Once we have 1) and 2), we need efficient AI/ML/Data Science Modeling and Algorithms which will do the magic (HOW part) of bringing Data Sources and Intelligence to gather. I have project needs.. At minimum, I can settle my needs with Kelly;s Blue Book Price Price Estimator API access.. Cheapest.. I posted this project becasuse I wanted comprehensive solution with some my data, some rules, some scraping and some APIs like KBB Phase 1- Technical Feasibility/ Data Sources/ technical solution Proposal - Identify what you can do with all approaches, factors that can be taken care of -- No development here -- I will pay for research (token) Phase 2 -- Prototype Demo = Use real data, API integration, fabricated data, etc.. - Rough Phase 3 - Final Production Ready API based solution(put "car" into your bid) 1) technical feasibility/ technical documentation / identify data sources- APIs and regression or Ai models and clear proposal - no documentation 2) prototype development based on proposal - bugs are ok - trial data integrations - some data manufactured and show our solution is paractlcal and indeed AI driven covering many factors of selling price 3) develop bug proof solution wirh all data integrations / production quality and ApI driven Check attached document for full explanation. Get back to me with your detailed proposal. Thanks.