Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/maxtheman/ai_estimator

ai_estimator
https://github.com/maxtheman/ai_estimator

Last synced: about 2 months ago
JSON representation

ai_estimator

Awesome Lists containing this project

README 

        
# ai_estimator



This file will become your README and also the index of your

documentation.



## Install



\#AI Estimator



This code is a proof of concept intended larger project that aims to

assist with property management tasks. It is designed to help with the

estimation of materials needed for a roofing job. It takes into account

various factors such as the size and shape of the roof, the type of

materials being used, and the specific requirements of the job.



## Calculations



The code uses the OpenAI API to interact with the user and perform tasks

based on the user’s input. It can create estimates for property

renovations or repairs, retrieve property owner details from the

database, retrieve property information from the database, retrieve

building structure information from the database, and retrieve materials

templates from the database.



The code is organized into several classes and functions, each with a

specific role. For example, the

[`SupportedTrade`](https://maxtheman.github.io/ai_estimator/core.html#supportedtrade)

and

[`RequiredRoofingMaterials`](https://maxtheman.github.io/ai_estimator/core.html#requiredroofingmaterials)

classes define the types of jobs and materials that the calculator can

handle. The

[`Material`](https://maxtheman.github.io/ai_estimator/core.html#material)

class represents a specific type of material, including its name, role,

how many units are in a package, the cost per unit, and a waste factor

to account for any material that might be wasted during the job.



The

[`RoofInputMeasurements`](https://maxtheman.github.io/ai_estimator/core.html#roofinputmeasurements)

class represents the measurements of the roof, including the length and

width of the area, the total length of the roof area, and specific

measurements for different parts of the roof like ridges, valleys,

rakes, etc. The

[`RoofingSettings`](https://maxtheman.github.io/ai_estimator/core.html#roofingsettings)

class represents the assumptions for a roofing calculator, like how many

shingles are needed per square foot of roof, how many nails are needed

per square foot, etc.



The

[`MaterialTemplate`](https://maxtheman.github.io/ai_estimator/core.html#materialtemplate)

class represents a template for a specific brand of materials, including

the name of the template, a list of materials, and the roofing settings.

The

[`LineItem`](https://maxtheman.github.io/ai_estimator/core.html#lineitem)

class represents a line item in the estimate, including the material

needed, the number of packages needed, and the cost per package. The

[`MaterialList`](https://maxtheman.github.io/ai_estimator/core.html#materiallist)

class represents a list of materials needed for a roofing job, including

the trade (roofing) and a list of line items.



Finally, the

[`RoofEstimate`](https://maxtheman.github.io/ai_estimator/core.html#roofestimate)

class is the main part of the calculator. It takes in the required roles

of materials, the measurements of the roof, and the material templates.

It then validates the templates, calculates the quantity and cost of

each material, and generates an estimate of the materials needed for the

job.



\##AI Usage



Uses Langchain’s `ChatOpenAI` Agent and tool decorators to provide the

above calculations to the Agent. Examples are provided showing how the

data is supposed to flow and a hardcoded example given a single full

property.



## How to to install



1.  Clone this repo

2.  Set `OPEN_API_KEY`

3.  pip install requirements.txt

4.  Then, open in VSCode with the Jupyter extension or start jupyter

    from command line. Development was done with `nbdev`



How to use:



First,



Run this in the terminal to run on Modal: www.modal.com:



First, `modal token new` to set up your own modal account



`nbdev_export` to get the latest package exported from jupyter

`modal serve ai_estimator/server.py` to see the docal preview



`modal deploy` to set it to run on your account



See demo running at:

https://maxtheman–ai-estimator-run-server.modal.run/