Machine learning platform for appearal
Budget: $1,500 – $3,000 USD
We want to create a platform for appearal where visitors can search appearal from their favorit influencers on Instagram.
Platform part
- Connect vendors with Channable
- Import all items which are in stock and use a link to go directly to the items in the connected webshop and let the customer finish the purchase there and with an affiliate link/pixel or similar calculate the commission of that particular purchase using an API in the currency of the vendor.
- Accounts for buyers, instagram influencers and vendors must be included
- Further all the common functions of a platform webshop with filtering on price, country, color, size, brand, etc.
- The platform needs to have Google Translate in order to get global coverage
The vendors
The vendors can sign up and connect their website witch Channable. For each vendor we must set a commission rate.
The buyers
The buyers can sign up and add their instagram account on the platform in order to select the Instagram influencer of their choice to search their clothing style and be able to find it and order it. The buyers must be able to select their fashion style in order to get recommendations from the platform. This can be done by ML based upon the last 5 purchases.
Instagram influencers
The Instagram influencers can sign up on the platform and will show on the platform. In case an Instagram influencer is not known on the platform an invitation must be send in order to sign them up and earn a commission. The commission rate can be set per influencer. Each influencer must authorise the use of their Instagram account in order to place advertising on their account.
The scan part
- Import images from selected Instagram influencer accounts
- Scan those imported images for the appearal (shirt, pants, shoes, handbag, accessories) with image recognition
The selection part
When an Instagram image is selected then a selection can be made (shirt, pants, etc.) and then the available items must show and when clicked on to be forwarded to the connected webshop. Ofcourse filtering on price, location, etc. must be included.
Milestones split in 10 sections:
1. ML/AI engine development (Amazon Rekognition)
a. Image comparing engine
b. Recommendation engine
2. Build backend for shopper
a. Account
b. Service area
c. Set parameters for recommendation
3. Build backend for Instagram influencers
a. Account with all credentials
b. Overview of earned commissions
c. Set preferred vendor (in case of sponsored deal so that particular vendor items will show!)
d. Import images from selected Instagram influencers by shopper for the Image Comparing engine
4. Build backend for vendor webshops (API or Channable-like connection)
a. Account
b. Generate tracking pixel for conversion (to be set on the order confirmation page of vendor)
c. Overview of commission rate/invoices
d. Service area (which country goods can be shipped to)
5. Build backend for Quest4Fashion
a. Access accounts of shoppers, Instagram Influencers, vendors
b. Invoicing and administration
c. API connection settings per vendor (based upon selection of standard records)
d. Shut down accounts of shoppers, Instagram influencers, vendors
6. Build front end for shopper
a. Shopping with filtering
b. Integration of Instagram (image selection, likes and/or comments)
c. AI/ML recommendation based upon previous visits so the shop opens instead of frontpage.
7. General
a. Design must be fashion.
b. Refresh rate of API connection to be optimized (every hour during working hours)
c. Every item that is coming from the API connection to be auto-categorized with image comparing engine (e.g. pants in pants, shorts in shorts, dress in dresses, detection of color, etc.)
d. Focus on men, women kids and babies
e. Additional categories: Lingerie, pajamas, hats, sunglasses and more to come.
f. API connection with ActiveCampaign
g. RSS feed
8. Testing
9. Bug fixing
10. Ready to market
If you have any questions, please let us know!
Platform part
- Connect vendors with Channable
- Import all items which are in stock and use a link to go directly to the items in the connected webshop and let the customer finish the purchase there and with an affiliate link/pixel or similar calculate the commission of that particular purchase using an API in the currency of the vendor.
- Accounts for buyers, instagram influencers and vendors must be included
- Further all the common functions of a platform webshop with filtering on price, country, color, size, brand, etc.
- The platform needs to have Google Translate in order to get global coverage
The vendors
The vendors can sign up and connect their website witch Channable. For each vendor we must set a commission rate.
The buyers
The buyers can sign up and add their instagram account on the platform in order to select the Instagram influencer of their choice to search their clothing style and be able to find it and order it. The buyers must be able to select their fashion style in order to get recommendations from the platform. This can be done by ML based upon the last 5 purchases.
Instagram influencers
The Instagram influencers can sign up on the platform and will show on the platform. In case an Instagram influencer is not known on the platform an invitation must be send in order to sign them up and earn a commission. The commission rate can be set per influencer. Each influencer must authorise the use of their Instagram account in order to place advertising on their account.
The scan part
- Import images from selected Instagram influencer accounts
- Scan those imported images for the appearal (shirt, pants, shoes, handbag, accessories) with image recognition
The selection part
When an Instagram image is selected then a selection can be made (shirt, pants, etc.) and then the available items must show and when clicked on to be forwarded to the connected webshop. Ofcourse filtering on price, location, etc. must be included.
Milestones split in 10 sections:
1. ML/AI engine development (Amazon Rekognition)
a. Image comparing engine
b. Recommendation engine
2. Build backend for shopper
a. Account
b. Service area
c. Set parameters for recommendation
3. Build backend for Instagram influencers
a. Account with all credentials
b. Overview of earned commissions
c. Set preferred vendor (in case of sponsored deal so that particular vendor items will show!)
d. Import images from selected Instagram influencers by shopper for the Image Comparing engine
4. Build backend for vendor webshops (API or Channable-like connection)
a. Account
b. Generate tracking pixel for conversion (to be set on the order confirmation page of vendor)
c. Overview of commission rate/invoices
d. Service area (which country goods can be shipped to)
5. Build backend for Quest4Fashion
a. Access accounts of shoppers, Instagram Influencers, vendors
b. Invoicing and administration
c. API connection settings per vendor (based upon selection of standard records)
d. Shut down accounts of shoppers, Instagram influencers, vendors
6. Build front end for shopper
a. Shopping with filtering
b. Integration of Instagram (image selection, likes and/or comments)
c. AI/ML recommendation based upon previous visits so the shop opens instead of frontpage.
7. General
a. Design must be fashion.
b. Refresh rate of API connection to be optimized (every hour during working hours)
c. Every item that is coming from the API connection to be auto-categorized with image comparing engine (e.g. pants in pants, shorts in shorts, dress in dresses, detection of color, etc.)
d. Focus on men, women kids and babies
e. Additional categories: Lingerie, pajamas, hats, sunglasses and more to come.
f. API connection with ActiveCampaign
g. RSS feed
8. Testing
9. Bug fixing
10. Ready to market
If you have any questions, please let us know!