Clari

A Pre-sales Management and Sales forecasting tool

Comment

Inspiration

  • Sales forecasting is the process of estimating how much product or service a company can expect to sell in the future.
  • It's important for companies to have realistic forecasts in order to stay on track with their goals and avoid surprises that could negatively affect their business growth. Companies want to know how much demand they'll have in the coming months or years, but predicting future demand is always difficult. But what if there was a better way?
  • Hence, I bring solution as Clari-- A Pre-sales Management and Sales forecasting tool.

What it does

  • Clari is sales forecasting tool that has been transformed by new technologies such as machine learning. This has improved the accuracy of predictions by making them more precise and accurate.
  • Clari is a Pre-sales Management and Sales forecasting application that allows a business to estimate future sales for a specific timeframe.
  • Clari enable companies to predict future growth trends and help leadership formulate effective strategies to expand their business.
  • Clari can predict number of item sales per month and that enables companies to prepare for future possibilities.

How I built it

  • This project is about predicting total sales for every product and store of a large Russian software firm, 1C company using Machine Learning techniques. The company has provided a challenging time-series dataset consisting of daily sales data in 6 different csv file.
  • I built this tool using frontend tech such as HTML, CSS and JavaScript and backend as Python and its library including matplotlib, Statsmodels, NumPy, Pandas, seaborn.
  • First of all I Understanded the story of my data for which I performed Exploratory Data Analysis.
  • Subsequently, I learned about traditional Time series analysis technique and its models such as AR, MR and SARIMA.
  • During the TSA workflow, I checked the Data Types of Time Series and must accessed if the given dataset is Stationary or not using Augmented Dickey Fuller Test.
  • Finally prepared a SARIMAX Model and derived a graph that shows the original and predicted value for sales.

Accomplishments that I'm proud of

  • The RMSE score was 18294 which shows that the model could relatively predict the data Accurately.

Built With

Share this project:

Updates