1
00:00:00,250 --> 00:00:03,602
A shopper has fifteen hundred rupees and already owns cleanser.

2
00:00:03,602 --> 00:00:05,388
What should they buy next?

3
00:00:05,388 --> 00:00:09,460
Goodskin combines a YouCam skin check-in with a routine planner that starts

4
00:00:09,460 --> 00:00:11,246
with what they already own.

5
00:00:12,196 --> 00:00:14,647
Here is the actual working app.

6
00:00:14,647 --> 00:00:19,292
I select the built-in synthetic face, then explicitly consent before
starting analysis.

7
00:00:19,292 --> 00:00:22,474
This public demonstration uses no real person's face.

8
00:00:22,474 --> 00:00:27,229
Goodskin sends the image to YouCam and does not save a photo

9
00:00:27,229 --> 00:00:28,217
or profile.

10
00:00:29,167 --> 00:00:34,241
The server requests a signed upload, uploads the image, starts the Skin
Analysis

11
00:00:34,241 --> 00:00:36,467
task, and polls for completion.

12
00:00:36,467 --> 00:00:39,082
The credential stays on the server.

13
00:00:39,082 --> 00:00:44,430
This fresh response returns an overall score of seventy-nine and five
cosmetic signals.

14
00:00:44,430 --> 00:00:48,217
These are observations about one image, not a diagnosis.

15
00:00:49,167 --> 00:00:51,187
I choose Comfort and hydration.

16
00:00:51,187 --> 00:00:54,980
The returned moisture score of seventy-two is shown as context.

17
00:00:54,980 --> 00:00:59,588
The shopper stays in control: this score does not decide which product to

18
00:00:59,588 --> 00:01:00,190
buy.

19
00:01:00,190 --> 00:01:02,211
Their preferences and budget do.

20
00:01:03,161 --> 00:01:08,043
Now I mark cleanser as already owned and choose a lightweight texture.

21
00:01:08,043 --> 00:01:13,040
With the fragrance-free filter on, the planner compares matching
combinations in our seven-item

22
00:01:13,040 --> 00:01:14,078
sample catalog.

23
00:01:14,078 --> 00:01:17,807
Gel and sunscreen total eleven hundred and forty-eight rupees.

24
00:01:17,807 --> 00:01:23,074
That leaves three hundred and fifty-two rupees, with no duplicate cleanser
to buy.

25
00:01:24,024 --> 00:01:27,417
What if the budget is only eleven hundred rupees?

26
00:01:27,417 --> 00:01:31,160
The app shows a forty-eight rupee gap and disables Add.

27
00:01:31,160 --> 00:01:35,708
It explains that this is the cheapest matching combination, so the shopper
can

28
00:01:35,708 --> 00:01:38,402
change preferences or reconsider the missing steps.

29
00:01:39,352 --> 00:01:43,054
I restore the budget and add only the missing steps.

30
00:01:43,054 --> 00:01:47,448
The bag contains gel and sunscreen for eleven hundred and forty-eight
rupees.

31
00:01:47,448 --> 00:01:49,074
The cleanser stays out.

32
00:01:49,074 --> 00:01:53,123
This is a working prototype bag, with no checkout or orders.

33
00:01:54,073 --> 00:01:58,958
Removing the gel brings the subtotal to six hundred and ninety-nine rupees.

34
00:01:58,958 --> 00:02:03,959
Adding the missing steps again restores the gel without duplicating the
sunscreen already

35
00:02:03,959 --> 00:02:05,382
in the bag.

36
00:02:05,382 --> 00:02:08,344
The interaction keeps the purchase decision reversible.

37
00:02:09,294 --> 00:02:12,862
Each item explains its price and preference match.

38
00:02:12,862 --> 00:02:17,660
A downloadable decision card is available, without the photo or credential.

39
00:02:17,660 --> 00:02:22,991
All product names, prices, and attributes here are illustrative; this is a
catalog

40
00:02:22,991 --> 00:02:23,688
sandbox.

41
00:02:24,638 --> 00:02:29,377
Our proposal for skincare retail is simple: understand the context, reuse
what you

42
00:02:29,377 --> 00:02:33,278
have, and make the cost of the next step explicit.

43
00:02:33,278 --> 00:02:36,814
The integration works; the planner has eight passing tests.

44
00:02:36,814 --> 00:02:39,986
Real shopper feedback is our next validation step.