web portal service for ai analysis of coffee roast


# Coffee Roaster Log Analysis API service and Web Portal

## Provide a portal to upload coffee roaster log files which will:

 - use the API below to convert log file to analysis package.
 - database the results along wtih log file.
 - display the roast session info, with summary and thumbprint image.
 - mouse-over or click for detail view of that roast session.
 - ability to group sessions together into single context (AI only looks at last 3-5?).
 - ability to plot metrics on site.


## Provide a REST API for a service that will:

Parse the coffee roaster log file and provide analysis results.

 - support multiple roaster log formats, dont assume one log file format.
 - read Artisan Scope "alog" files, parse with 'ast' python dict archive files.
 - Other roaster logs format??

Parse commonly used key metrics in every roast:

 - coffee variety, batch size
 - ambient temperature, humidity.
 - record metrics like gram before/after weight loss, bean density
 - DTR ratio, TP, DE, FC, DROP.
 - Cooling curve exponential ratio (k, Y0, Tau, half-life, ...)
 - temperature scale C or F.
 - return these in 'metrics' section of json data model.


### Use artisan in cli (or make cli interface?) to:

 - generate some chart img (png) - one generic, one with stats overlayed

### Prompt claude.ai for analysis

 - use a long chat context per portal account (per beans type and batch size?)
 - allow grouping in portal for AI context.
 - keep the AI response with this roast, along with 'summary' from AI.

Use the following prompt:

```
Analyze this attached Artisan Roasting Scope .alog file. Provide analysis
of the roast plan used, and provide guidance for improvements for these
same green coffee beans and same batch size using the same 500 gram
electric drum roaster. The temperatures listed are in C not F.
```

or with metrics:


```

Analyze this attached Artisan Roasting Scope .alog file. Provide analysis
of the roast plan used, and provide guidance for improvements for these
same green coffee beans and same batch size. Organize the analysis
into sections titled:
 "Profile Analysis",
 "Issues Identified",
 "Specific Problems vs. Roasting Notes",
 "Recommendations for Next Roast",
 "Expected Improvements",
 "Roast Summary".
Under the section "Roast Summary" provide a short summary and top
three bullet point items for this file.
The coffee bean roaster is a 500 gram electric drum roaster.
The temperatures listed are in C not F.
Refer to "the roaster" in third person instead of "you" second person.

```



##