Project success prediction

• 

  ML process

Inspiration

We have been inspired by day to day work on customer implementation projects. There is a lack of predictability for a project status in different project phases.

What it does

The pilot helps to predict when project becomes "red", based on the historical data of defects from previous projects, and suggests potential actions to prevent the crisis.

How I built it

Provide feed from the following sources to ML: from Jira with extraction of defects' backlog from closed and live projects. PowerBI to provide a dashboard with project statuses WIX to present the idea and flow

Challenges I ran into

Data collection process and how to define indicators for the model creation

Accomplishments that I'm proud of

TBD

What I learned

TBD

What's next for Project success prediction

Model extension to take into account the following parameters/inputs/indicators: 1.Prioritization of customers according to customer profile/ranking

1. Use test packs and automation ratio of Finastra and customers
2. Use OpenAir resource time allocation
3. Use Salesforce escalation indications

Built With

• jira
• ml
• openair
• powerbi
• wix