Krishna N. Data Engineer

Experience

Data Scientist III at Walmart Global Tech, Bangalore, India

07/2022 – today

Product: Walmart Luminate - B2B supplier support app used by multinationals like
Nestlé, Procter & Gamble, Johnson & Johnson for sales data analytics.

• As part of the solution, build an anomaly alert ranking module to detect, prioritize
  and contextualize anomalies for Walmart products.
• Collaborate with the data engineering team to develop and implement best coding
  practices like code modularity and productionisation.

Technologies: Anomaly recommendation | Statistical modeling | Regression | Machine
Learning | Time series analysis | Pyspark | SQL | Element | Git

Senior Data Scientist at Bosch Global Software Technologies, Bangalore, India

04/2020 – 06/2022 
Project: ITAM - Intelligent Tyre Asset Management System for CEAT (leading tyre
manufacturer) to check real-time tyre health and performance.

• In charge of the end-to-end project development life-cycle, leading the analytics
  team of four data scientists. Built predictive models using machine learning algorithms (XGBoost, Regression analysis), statistical modeling, and digital signal
  processing techniques to improve the predictability of tyre performance and reduce
  costs incurred in managing tyres.
• Oversaw engineering of the collected data to directly read from sensor logs to a
  model-ready format. Developed an app using Python, and regex to perform the data manipulation task that works well for various sensor versions. This has
  significantly saved hours of effort to perform manual data manipulation for field
  testers. Visualized signals, analyzed data, extracted features, built models,
  proactively suggested new use cases, and presented insights to CEAT that
  improved overall safety and downtime by tyre accidents. Received Bosch Agile
  Award for project excellence.

Technologies: Machine Learning | Signal Processing techniques (FFT, Autocorrelation,
Wavelet transform, PSD, Filtering) | Statistical modeling | Pattern Recognition | Regex |
XGBoost | Regression | Python | Lime | Scikit-Learn | Pandas
Product: Phantom Edge - an AIoT platform with real-time views of electrical energy
consumption and machine-level information used by German industry clients like
KERN-LIEBERS and Milltec.

• In charge of automating cycle time detection and signal pattern identification
  modules using ML, Statistical Modeling, and Signal Processing Techniques for the
  2.0 version.
• Automated the cycle time capture from signals and measured accurate downtimes
  by providing timely, bias-free, and precise data that form the basis for managers to
  set targets, track performance, analyze, and improve continuously. The algorithm
  built was fully generalized to work well for all small and medium-scale machines
  without any manual intervention. Worked for 40+ clients and took part in their
  industry 4.0 digital transformation journey. Received Extra miler award for team
  contributions by Bosch.

Technologies: Machine Learning | Signal processing techniques (Fourier Transform, Wavelet scattering, Filtering, Smoothening) | Feature extraction | Statistical modeling |
PCA | Pattern matching | Clustering analysis | Time series analysis | Correlation |
Anomaly detection | Python | Scikit-Learn | Pandas | NumPy.

Data Analyst at EY Global Delivery Services India LLP, Trivandrum

04/2019 – 03/2020 
Project: Management services bot - Responsive & AI-based chatbot for customer
interaction with Guidewire (US insurance company) for customer queries (text/voice).
Played a key role in development/integration/deployment. Trained Microsoft LUIS for
intent classification with customer data and leveraged phonetic algorithms to
overcome errors from misunderstanding of voice input. Built and integrated a
D3.js-based visualization module for providing graphical responses for customer
queries on coverage/plan. Integrated Azure speech-to-text and text-to-speech APIs
with the application.
Technologies: Python | Flask | REST API | Microsoft SQL Server |
Microsoft LUIS | AIML | Phonetic algorithms | Azure speech services | D3.js

Project: Policy bot - Responsive chatbot for user queries on insurance policies.
Designed the bot, a pure Python-based search engine to match user queries with the
policy PDFs and deployed the final solution in Azure.
Technologies: Python | Flask | Whoosh | Tika | MongoDB | Azure app services | Git.

 Machine Learning Developer at TATA Consultancy Services Limited, Coimbatore

11/2016 – 04/2019
Project: Gilbarco Chatbot - AI engine developed for Gilbarco Veeder-Root (US fueling
equipment supplier) to monitor the performance of fuel dispensers/meters across fuel
stations in real-time. Developed a hybrid model from scratch that predicts working
conditions, next week's flow rate, and the survival probability of the fuel dispensers
(clog/normal) using machine learning algorithms and statistical analysis. Worked with
2 developers so they translated the model to the bot. Reduced downtime for
maintenance because of a data-driven decision to optimize maintenance planning.
Technologies: Python | Trend analysis | Binary classification | MLP | Survival analysis |
Regression analysis | SQL.

Project: Auto-categorization of IT support tickets - Built ML algorithm for app service
providers to manage/automatically resolve support tickets. Used a hybrid approach
that uses both unsupervised clustering and supervised deep learning embedding to
maximize the feature exploration and learning efficacy. Multiple optimized models are
assembled to build a recommendation engine - reducing the time to resolve the tickets
without multiple reopening/long triaging by 50%.
Technologies: Python | Topic modelling | NMF | LDA | NLP | Clustering | LSTM | Keras | Deep neural networks | Tf-idf.

Education

Bachelor of Technology in Electronics and Communications Engineering, N S S College of Engineering, Palakkad, India
08/2012 – 07/2016

Certificates

• Microsoft Certified: Azure AI Fundamentals
• Applied Data Science with Python

Achievements

• Winner of Independence Day medical imaging hackathon (Bosch)
• Extra miler award for team contributions (Bosch)
• Bravo award for mentorship (Bosch)
• Agile award for project excellence (Bosch)
• Part of ”PHANTOM IoT Solu- tion” that won NASSCOM Innovation Award for “Engin- eered in India Product of the Year 2021” & finalist in ”Bosch India Innovation award 2020”

Patents

An apparatus and method to estimate cycle time for processing
a workpiece in an industry - Application number: 202141034096, Jul 2021, filed
together with Bosch, status pending with Indian Patent Office.

A system and method to monitor the operation of a machine in a
manufacturing industry - Application number: 202141034093, Jul 2021, filed
together with Bosch, status pending with Indian Patent Office.