[ad_1]
I’m a Information Science skilled having 13 years of expertise within the IT trade. I’m at present working as a Sr. Information Scientist with Autodesk. Earlier than becoming a member of the PGP-AIML program, I used to be a Information Science Skilled.
The shopper help crew at Autodesk receives many utility errors stories whereby the crew manually verifies these stories and routes them to the respective groups/organizations to unravel the problems. That is an unsupervised state of affairs the place we now have utilized the ML method to learn by way of the context of the error and categorize the error information (Buyer Error Report)
Each utility/product of Autodesk generates many Buyer Error Stories ( these are the error log recordsdata), that are saved within the AWS cloud platform. Our Information Science crew extracts these stories( textual content recordsdata), extracts required information from these error stories, cleans the information, applies textual content information cleansing strategies, and makes use of the Okay-means Clustering algorithm to cluster the stories.
This delays fixing the shopper situation and requires guide effort to route the problem to the correct crew.
Now we have used the Python software NLTK to scrub the textual content information and the Okay-Means clustering algorithm to cluster the stories. And AWS lambda to deploy the mannequin.
Over the interval within the AI/ML course, I realized many strategies to unravel a few of the difficult enterprise issues. So I discovered the chance to use ML strategies like clustering/ classification (if we now have supervised information) on this specific enterprise drawback to unravel the problem, and it occurred.
Now we have really helpful the answer to cluster the error stories in several segments and route them to the correct crew shortly to take instant motion.
After implementing this resolution, we now have elevated buyer effectivity and decreased the turnaround time to unravel the error situation. Additionally lowered manpower which saves 3FTE yearly.
This was a terrific mission I labored on the place I’ve carried out my learnings from the AIML course. Nonetheless, there are numerous issues to study, and I shall be utilizing my studying to unravel many extra enterprise issues.
[ad_2]
