United States

Data Scientist

Data Analysis

Bachelor's degree

Master's degree

Sijo Valayakkad Manikandan

Lead Data Scientist

Collaboration

Problem-solving

Time management

Stakeholder management

Advanced data analytics

Data mining

Machine learning

Pandas

Python

Numpy

» Leading end-to-end data science projects, empowering commercial analytics and supply chain optimization

» Led the development and implementation of a multivariate demand forecasting solution for Nike Stores in North America 

» Developed and implemented probabilistic forecasting models using Quantile Regression and ETS, resulting in a transformative impact on the monthly Merchandising Financial Planning Process

Quantiphi is a category-defining Applied AI and Machine Learning software and services company focused on helping organizations translate the big promise of Big Data & Machine Learning technologies into quantifiable business impact.

» Developed demand forecasting models including exponential smoothing, seasonal naive, and moving average along with stacking regression models to forecast inventory level data - achieved the organizational goal of <20% sMAPE resulting in a $15M savings over the next 5 years for a leading fast-food chain

» Lead a team of 4 ML practitioners to automate forecasting experimentation pipelines using Apache Airflow, transitioning to an agile experimentation process

» Developed an improved feature engineering pipeline for the XGBoost model to build a Fraud Detection Solution for a financial services client, outperformed the existing AUC by 6%, and secured an 850K USD engagement

» Built a Vehicle Damage Cost Estimation Android Application for Auto-Insurers using pre-trained image classification models and car parts API for GTM activities, selected for presentation at NVIDIA GPU Technology Conference 2022

CVS Health is the world's largest publicly traded healthcare company.

» Analyzed the underwriting data and created new rule-based KPIs to standardize metrics across the organization – instrumental in developing a Rules Hub, a product to be used by the Underwriting Analytics team

» Automated the weekly carrier setup process using python – ensuring data quality and reduced the processing time to 30 minutes from 4 hours

» Identified issues in one-day delivery orders in Ohio by developing Gradient Boosted Decision Trees and retrospectively analyzing shipment and order data, potential to uplift long-term revenue by around $1M USD

» Kickstarted the Machine Learning-based delivery date prediction project with satisfactory baseline models using ML techniques, with an MAE of 8.8 days (compared to the current MAE of 15 days)

» Led the development of a Data Quality Assessment System using a Confusion Matrix-based approach to measure and improve the data quality of the datasets to over 95% and enrich the annotation effectiveness in a Gesture Recognition Project for a luxury car manufacturing client

» Revamped data pipelines using python and SQL to preprocess 5-10 TB of image data every week and generate training-ready datasets for a gesture recognition project, reducing the dataset generation time by 33% compared to the previous version

» Analyzed data from different project tracking tools of multiple software testing projects and designed a visualization dashboard using Microsoft Power BI to generate a consolidated view of a client’s account - facilitated quicker resource allocation, efficient software release, and reduction in potential bottlenecks in the projects

» Built an automated Load Forecasting training pipeline using SPSS modeler and python to project medium-term energy demand for an energy client, assisting in load management and futures market contracts decisions

Sijo Valayakkad Manikandan🇺🇸

About Sijo Valayakkad Manikandan

Expertise

What mentees say