Ayush Chauhan

As a Data Science consultant, I have been involved in various projects across different domains, providing valuable data engineering and analytical solutions to my clients. Here are some highlights of my work:

Client Domain: FMCG Industry

• Successfully executed a data engineering solution on Azure, developing multiple azure pipelines to ingest sales data from diverse sources. I established a robust warehouse architecture to maintain incremental and historical data, enabling data consolidation for modeling and dashboarding purposes in the FMCG Industry.
• Conducted sales forecasting for various products in the FMCG Industry, leveraging four years of historical sales data and aiming to predict sales for the upcoming six months. By preprocessing the weekly sales data and generating relevant features, I employed advanced Multivariate forecasting models such as Prophet and LSTM. The accuracy achieved across all products was an impressive 90%.

Client Domain: Automobile Industry

• Developed scripts to perform a variety of statistical analyses, effectively helping clients to enhance product yield.
• Created scripts to generate daily plots, which enabled clients to monitor machinery and daily processes effectively.
• By identifying and addressing inefficiencies promptly, the daily production downtime reduced to 4 %, resulting in a production of 150 tire per day on average & subsequently potential revenue boost of 10.9 Cr /year.
• Utilized Machine learning models to predict precise parameters impacting tire manufacturing, leading to the creation of a superior tire batch.
• Implemented descriptive and prescriptive analytics to address frequent manufacturing issues, including overproduction, logistics, and inventory challenges.
• Collaborated with McKinsey & Co. to build a vendor/distributor churn model, predicting soft churn among vendors. This model serves as a valuable tool for the marketing team to understand vendor dynamics better.
• Developed a Statistical model to predict the failure of the extruder machine, a critical tire component manufacturing equipment. This predictive capability enables the plant to take proactive measures, preventing accidents and reducing downtime in the overall manufacturing process.

Through my expertise in data collection, ETL process development, statistical analysis, and machine learning models, I have consistently delivered high-quality solutions that meet the unique needs of each client in the FMCG and Automobile industries. My focus remains on delivering accurate and actionable insights to support data-driven decision-making and drive business success.

• Design and implement the management, monitoring, security, and privacy of data using the full stack of Azure data services to satisfy business needs
• Spearheaded the development of robust Data Pipelines incorporating diverse activities and seamlessly integrating data flow from various Azure and on-premises sources to meet the unique business requirements
• Demonstrated expertise in implementing an innovative Azure Data Factory (ADF) solution, enabling efficient Exception Handling, Error Messages, and Activity Tracking for every pipeline, ensuring streamlined operations
• Pioneered the creation of sophisticated Databricks SQL Analytics and Dashboard, providing real-time insights and comprehensive visualization of critical data metrics, elevating business decision-making to unprecedented levels
• Successfully designed and deployed a state-of-the-art Price Optimizer, leveraging live Streaming data and Change Data Capture (CDC) capabilities in Databricks, resulting in significant profit gains and unmatched accuracy in predictive models
• Revolutionized data ingestion processes by automating the extraction and transformation of daily-changing data with Azure Data Factory (ADF), enabling the generation of powerful insights on schedule, leading to substantial business growth
• Leveraged advanced Machine Learning algorithms in Databricks to develop predictive models that generated actionable intelligence, which was translated into visually impactful Power BI reports, empowering stakeholders with actionable insights.

• Encompasses the design, development, and execution of experiments for the assessment and enhancement of cognitive abilities through various ML/DL algorithms
• Create an emotional stimulus experiment in Python for subjects to perform the physiological test
• Worked on the data from EEG signals Deep learning Models (CNNs, RNNs, LSTM, and GRU & Bidirectional)
• RNNs with different architectures have been experimented with
• Assisted 3 scientists with in-depth data research in both lab and office environments
• Collect Data for Cognitive State Assessment Using EEG Signals &create a classifier
• Created an Engagement Detection System, which calculates the concentration index based on the facial angles, and pupil movement including dilation & facial emotions
• Researched detecting Mental Workload & fatigue through facial thermography.