Biswapratap Chatterjee

Data Architect and ML Engineer

Bangalore, India

Summary

Data Architect with 11 years of experience in machine learning, data pipelining, data

analytics/transformation and storage technology domain. Adept in designing and implementing

end-to-end data science projects. Loaded with a wide range of skill sets scaling from system

programming, protocol layer development, middleware development, infrastructure bring-up and

some presentation layer development. With innovation being the key strength and execution

equally complimenting can handle projects from start to finish.

Languages:

English

Favorite Python Packages:

sklearn, keras, tensorflow, pandas, numpy, flask

Experience

https://www.linkedin.com/in/biswapratap-chatterjee-42277518/

Data Architect with 11 years of experience in machine learning, data pipelining, data

analytics/transformation and storage technology domain. Adept in designing and implementing

end-to-end data science projects. Loaded with a wide range of skill sets scaling from system

programming, protocol layer development, middleware development, infrastructure bring-up and

some presentation layer development. With innovation being the key strength and execution

equally complimenting can handle projects from start to finish.

Experience

2017-03 -

present

Data Architect

Western Digital

QVisor

Jarvis

Optimaliseer

Log Analyzer

WD360- An Intra-Company Appathon Competition

An intelligent software/firmware quality analysis machine.

Targeted for validation testing of Western Digital firmware using attention based

reinforcement learning.

Patent applied, technical design.

The primary idea proposer, designer, architect and execution manager.

Responsible from POC developments for early risk mitigation to full execution

of the project.

An intelligent agent for managing different datasets produced by various teams

across Western Digital.

Automatic relationship determination, relationship heatmap generation, auto-

highlighting of hidden interesting data correlation, 2D and 3D graph projections.

Patent applied, technical design.

The primary idea proposer, designer, architect and developer.

A reinforcement learning agent aiding Western Digital Drive firmware to

enhance performance.

Auto tuning of firmware configurable parameters to achieve enhanced

performance.

Achieved 12% performance improvement as compared to human (expert)

tuning.

The primary idea proposer, designer, architect and developer.

An algorithm and machine learning based log analyzer for - Anomaly detection,

Parameter Extraction, Root Cause Analysis, Bug Classification.

Patent applied algorithm to auto cluster logs based on log structure and log

time.

Generate finite state machines from time based logs for debugging and root

cause analysis.

Log vectorization and clustering for bug classification.

A 360 degree contextual search engine for WD My Home (Personal Cloud).

The search engine could search based on n-grams, context and objects

(images/audio/doc).

Used Open CV, Sony Gracenote, IBM Watson and Elastic Search.

Won 1st prize worth $5000 in the competition.

2015-03 -

present

Personal R&D for skill development

Skills

Machine Learning – a) Python –

Keras, Keras-RL, Tensor Flow. b)

Supervised Learning. a. Classification

Problems. b. Decision making

Problems. c) Reinforcement Learning:

a. Q-learning. b. Policy Gradient

Learning. c. Deep Deterministic Policy

Gradient (Actor-Critic) Learning. d)

Genetic Algorithm e) Attention Based

Neural Networks f) GANs g)

Unsupervised Learning h) NLP i) RNN

j) LSTM

advanced

Data Analytics – a) Infrastructure. a.

Spark/Mesos b. Elastic Search. c.

MongoDB. d. Scalable REST APIs

using Falcon. b)

Algorithms/Techniques. a.

Pandas/Numpy/SciPy/SkLearn etc. b.

Web Scrapping. c. Natural Language

Processing (NLP). d. Image Analytics

using OpenCV. e. Audio Analytics

using Sony GraceNote. f.

Document/Text/Emotion Analytics

using IBM Watson. g. Density

Distribution. h. Timeline Distribution. i.

Predictive Analysis. j. Decision Tree

Classifiers.

advanced

Storage Application Development

and Sustenance. a) Enterprise and

Consumer Grade. b) SMIS, SNIA,

CIM, OpenSSL, OpenSLP. c) SCSI,

SATA, NVMe, PCI protocol

development stack. d) RAID and

Initiator Target Storage Controllers.

advanced

System Application Development

(Windows, Linux, VMware) – a) C/C++

b) RESTful Applications. c) Python. d)

Android. e) Java. f) NodeJS/Electron

advanced

Self

All the projects mentioned below are personal projects or online competitions

done in private time in parallel with professional work.

Case Resolver

Buddy

Problem Statement: A company has more than 50,000 clients and there is a

dedicated Field Application Team (FAT) for each client. The FATs and the clients

communicate via various means like chat, email, Jira/bug tool etc. The FATs

and the clients sometimes do not use grammatically correct English sentences

for their communication. Moreover the FAT often mentions the name of the

same client in various forms like - IBM, IBM Pvt. Ltd., IBM Corp, IBM Global etc.

The company wants to segregate all communications by unique client names

and also figure out the accurate date for a case to have resolved from the

discussions between the FATs and the clients.

Problem was solved using algorithms, supervised ML, Natural Language

Processing, Sentiment Analysis and Regular Expressions.

Although the problem was solved but the accuracy could still be improved.

An automatic e-shopping agent.

We spend a lot of time browsing over the internet in e-commerce websites over

millions of products on a daily basis. Still we are never sure if we made the best

choice, not just w.r.t. price but personal choice (which can be many).

This ML agent solves this problem of the user using - algorithms, web

scrapping, genetic algorithms etc.

Although the problem was solved but the accuracy could still be improved.

2016-09 -

2017-03

Senior Firmware Developer

Seagate

Dothill

This project was about the sustenance of the well-known product – Dothill

Storage Array. The storage array was used to address the mid-range of the array

market with various features involving SATA and SAS drives. The project required

me to work on the sustenance roadmap of the Dothill storage array. I worked in

the night US shift to address high priority customer issues by analysing Firmware

logs and providing recovery or code fix action plans. I was involved in multiple

modules of the product from – Platform, Backend, RAID, Cache, Paged Storage,

Snapshot and Tiering.

2015-06 -

2016-09

Senior Staff

Broadcom

Management Stack of 1G, 10G and 50G Network Controllers

The project involves the development and maintenance of the software

management stack for high end network controllers from Broadcom. I am acting

as a prime engineer in designing and implementing the entire stack. I have

complete knowledge on the stack starting from the CLI, CIM till the OS library.

2013-03 -

2015-06

Staff Engineer

Sandisk

Projects related to storage protocols like - SATA, SCSI, NVMe

2008-09 -

2013-03

Senior Software Engineer

Wipro

Projects related to storage protocols like - SATA, SCSI, NVMe, RAID

Web Application Development – a)

AngularJS 4+ b)

JavaScript/JQuery/TypeScript/HTML/

CSS

hands on

Search Engine Development.

average

Android Application Development

in Java.

hands on

Education

2004-05 -

2008-05

Kalinga Institute of Industrial Technology (KIIT) University

Bachelor of Technology in Computer Science

CGPA of 8.77

Skills

Artificial Intelligence (AI), Big Data, Data Science, Elasticsearch, Machine Learning, Natural Language Processing