K H M BURHAN UDDIN

A full-stack software engineer. A continuous learner who loves to work on the edges of the latest technologies.

  • Built first-ever Bengali search engine during undergrad, 1M funding from multinational telecom (Telenor).
  • Founding senior SWE at a high-growth AI startup, shipped products within 6 months.
  • Experienced in developing enterprise java applications, data engineering, and ETL data pipelines from scratch.
  • Worked with AWS cloud technologies and developed highly scalable microservices and APIs.
  • Profound understanding of algorithms, data structures, OOP, design patterns, and system design.

Experiences

Goava AB, Sweden

Dhaka, Bangladesh
  • Senior Software Engineer (Data Engineering, Founding SWE)
    April 2017 - Present
  • AWS
  • Deep Learning
  • Elastic Search
  • Java
  • Python
  • Spring Boot
  • Microservices
  • GraphQL
  • Data pipeline: Developed petabyte scale ETL pipeline for crawling and storing data from millions of sources. Designed architecture from scratch and led the project development (with 3 SWEs), served 1K customers within 6 months that helped the company secure EU grants.
  • Cloud crawler: Developed distributed crawler to crawl ten million websites every day.
  • ElasticSearch cluster: Developed and maintained the architecture for searching and indexing terabytes of data.
  • AWS cost optimization: Reduced AWS Cloud billing by about 50% by introducing batch data processing.
  • Recomendation system: Launched Goava Discover, a petabyte-scale recommender service for B2B sales.
  • Microservice and API development: Created scalable microservices and REST APIs for data pipeline.
  • Admin portal: Developed the full stack admin portal for automation of sales team using spring boot and react.

Nilavo Technologies Limited (Biscom Inc., Boston, U.S.A.)

Dhaka, Bangladesh
  • Senior Software Engineer
    August 2015 - March 2017
  • Software Engineer
    February 2013 - July 2015
  • Junior Software Engineer
    April 2011 - January 2013
  • Spring
  • Struts
  • Java
  • Java EE
  • Hibernate
  • JPA
  • JavaScript
  • AngularJS
  • MySQL
  • Postgres
  • REST
  • C#
  • ASP.NET
  • Major technology upgrade in enterprise application: Created web-based cross-browser multi-threaded file transfer library to send files larger than 100+gb. It replaced the old java applet-based library that only supported megabytes of non-resumable transfer.
  • Introduced resumable file transfer in encrypted data: Implemented file download pause/resume support for encrypted files by implementing random access in the java input stream by using block-by-block decryption of AES.
  • High availability file system design: Developed (lead the product development with 2 other junior SWE) a hierarchical file system using SQL and java to support high availability. Developed complete file system frontend tree explorer user interface from scratch using AngularJS.
  • Large file compression: Implemented GZ file compression in java to compress large files in the server.
  • Audit logging: Developed architecture of audit logging module with log versioning support using reflection.
  • Security layer architecture: Added proper permission check and validation in business logic using template pattern.

Skills

  • Languages: Java, Python, JavaScript, NodeJS, SQL, C#
  • Databases: Relational (MySQL, Postgres, MSSQL), NoSQL (MongoDB, DynamoDB)
  • Fundamentals: Algorithms and Data Structures, Design Patterns, OOP, Event-Driven Architecture
  • Cloud, Library & Frameworks: AWS, Spring Boot, Hibernate, ElasticSearch, Microservices, CQRS, Redis, Hadoop
  • Machine Learning: Deep learning, PyTorch, NumPy, Pandas, Tensorflow, Jupyter
  • API Development: Rest, GraphQL, gRPC, Web Services
  • Frontend: React, NextJS, AngularJS, BootStrap, HTML, CSS
  • DevOps: Docker, Git, Linux, Bash, CI-CD, Jenkins
Most Frequently used

Research

Pipilika Search Engine

December 2008 - April 2013
  • Hadoop
  • Lucene
  • Java
  • Java
  • Crawling
  • Indexing
  • Searching
  • Distributed Systems
  • Proposed and developed a prototype as an undergrad thesis for the first-ever Bangla search engine named ”Pipilika”.
  • Conducted R&D on web crawling, indexing of joint Bangla alphabets, and integrated distributed systems (Hadoop). Proposed and developed a generic content parser with 80% accuracy.
  • Pitched and demonstrated the product on several occasions and received several awards. Upon receiving 1M funding from Telenor, we launched it to production in 2013. Our work has been featured in several national news media.
  • Since then it has served as a data platform for researchers at SUST, leading to several NLP research publications.

Paperless Admission System

2009 - 2010
  • Automation
  • Digitalization
  • Automated and digitalized admission of SUST. For students, it reduced traveling across the country and filling redundant forms to simply sending an SMS. At university, it reduced error-prone paperwork from 30 days to less than a second.
  • As our work reduced carbon emissions, our project was inaugurated by the prime minister of Bangladesh herself. It revolutionized admission processing in the entire country as all the universities followed our model later.
  • SUST received funding for building a high-rise ICT complex from the education ministry of Bangladesh for this project.

Projects

Sentiment Analysis & Deployment

Created a sentiment analysis model using IMDB movie reviews and deployed the model in AWS SageMaker.

  • Deep Learning
  • AWS
  • NLP
  • SageMaker
  • PyTorch
Face generation

Used DCGAN to generate fake images of humans using celebrity face data.

  • Deep Learning
  • Generative Adversarial Network
  • PyTorch
Generate fake TV script

Created a model using a recurrent neural network to generate fake tv scripts.

  • Deep Learning
  • Recurrent Neural Network
  • PyTorch
Landmark classification and tagging

Used CNN to detect landmarks images then used transfer learning using ResNet.

  • Deep Learning
  • Convolutional Neural Network
  • PyTorch
Predicting bike sharing patterns

Created a neural network from scratch to predict bike sharing patterns.

  • Deep Learning
  • Neural Network
  • Pandas
  • Numpy

Education

Shahjalal University of Science and Technology (SUST)

Sylhet, Bangladesh
BS in Computer Science & Engineering
November 2006 - March 2011

Certifications and Tests

Deep Learning Nanodegree

May 2022
Udacity

Scalability & System Design for Developers

May 2022
Educative

IELTS

January 2020
7.0 (READING: 7.5, LISTENING: 6.5, WRITING: 6.5, SPEAKING: 7.5)

Awards

Pipilika Search Engine

National Collegiate Software Contest (NCSC) | 1st Prize
2011
Digital Innovative Fair | 1st Prize
2010

Paperless Admission System

mBillionth Award, South Asia
2010

Extra Curricular

CENTER FOR RESEARCH & CONSULTANCY (CRTC), SUST

ASSISTANT SOFTWARE PROGRAMMER
2007 - 2011

SUST OPEN SOURCE NETWORK, SUST

ASSISTANT GENERAL SECRETARY
2009 - 2010

CODESPRINT, BDOSN

PHONETIC BANGLA TYPING BROWSER PLUGIN
2008

WORKSHOPS CONDUCTED

OBJECT ORIENTED PROGRAMMING (JAVA)
2009 - 2010
WEB DEVELOPMENT (JAVASCRIPT, PHP)
2009 - 2010
OPEN SOURCE AND LINUX
2009 - 2010
Flash 2D Animation
2007 - 2008