FNU SHASHANK

Machine/Component Failure Detection and Recommendation :
OCI telemetry data analysis and predict the failure of engine for Premier OCI customer for its manufacturing facility by understanding the Telemetry data for future forecast and maintenance by the use of machine learning techniques.

• Applying ML techniques to provide actionable insights to improve quality and efficiency of customer support.
  Involved in end-to-end project development - requirement gathering, formulating hypothesis, testing hypothesis, development, deployment, and feedback loop.
  Using ML and word embedding for classification of textual data.Applying ML techniques to provide actionable insights to improve quality and efficiency of model building. Involved in end-to-end project development - requirement gathering, formulating hypothesis, testing hypothesis, development, deployment.
• Skills: Hypothesis Testing · Python (Programming Language).Machine Learning · SQL, Deep learning, Time series.

Bug text summarization for BugDB and Similarity of documents:
Use Natural language processing to categorize bugs belonging to the area of product development.
Enormous data present in the bug DB needs to be extracted and categorized based on the summary text
and also analysis on product development with the text summarization using Gen AI.

Implemented document parsing and clause extraction algorithms for input documents (PDF and DOCX).

Accelerated data labeling process by implementing techniques such as text clustering and active learning algorithms.

Improved the Gen AI models through prompt engineering,performance check through LORA.

Derived useful insights on clause negation and debugged errors in labeled data.

Objective:
- To emphasize the product dev teams about the important areas within the product where the bugs are created like performance, configuration.
- Bug description summarization through transformers.
- Helps the leadership to concentrate and prioritize the product development area with tech stacks to avoid regression in the development along with Similar document retrieval.

Skills used: Machine learning, Natural language processing( NLP),Gen AI, NLP/LLM model development, AI/ML pipelines, Python.

Customer Churn for OCI:
Oracle Cloud Infrastructure oracle's own platform which is onboarding customers across the domains in the industry.
The project focused on years worth of data for customer OCI usage metrics and predicting the churn for the customer.

Result:

Huge turnaround of 11% achieved for the customer success.

Customer Segmentation OCI:
For the OCI internal sales team , requirement was to get the customer segmentation of OCI usages.
Customer segmentation for the better approach and retention.

Skills Used: Python, Clustering, Machine Learning Algorithms,

Result :

Achieved 23 % conversion on the customer retention.

Sales Forecasting and Demand Planning :

This project was about developing framework to Forecast/Predict the Gross Sales Value (GSV) of different Products based on different
Predictor Variables and then plan for their demand for next month.

Skills Used: Python, Machine Learning Algorithms, Time series.

Client : VISA INC

• Designed and implemented machine learning system that predicts anomaly detection with more than 80% accuracy.
  •Created global and personalized real time reports system for executives stakeholders and processes in SAS, Tableau, and proprietary systems.
  •Worked closely with team of data engineers and BI analysts to improve efficiency customer recommendation analytics engine by 33%.
• Turned data into actionable insights, providing C-suite stakeholders with insightful recommendations to streamline business operations and improve customer experience.
  •Generated statistical reports and visualizations, providing key insights for more than 20 marketing campaigns and initiatives.

Data scientist :

Client: American Express.

Responsibilities Included:

Assisted in analyzing large sets of customer credit card data and providing insights to the marketing team.

•Assisted in pulling and analyzing large datasets, cutting operation time by 15%
•Contributed to marketing strategy by providing data-driven insights, boosting leads by 10%
•Managed database using SQL, which ensured data integrity and availability.

• Developed and implemented machine learning model that improved efficiency by 20%

Client : American Express

• Collected technical requirements for $500K+ customer accounts, defining data rules for and KPIs for performance metrics.
  •Perform HR data collection and variety of statistical analyses using Microsoft Excel, SAS, Tableau and Python.
  •Assisted senior data science team in building innovative machine learning models and segmentation for personalization initiatives to drive margin, revenue, and conversion.

Client : IBM

Responsibilities Included:

• Data visualization analysis for manufacturing nano chips.
• Building linear models and presenting them to leadership.
• Increased model accuracy by fine tuning for better results.

Client : IBM

Responsibilities included:

· Update/Fresh install of code in lower/Production environments.

· Fixing issues related to Deployment and engaging suitable application teams for same .