Ankush Malik
Last Updated on 23rd Feb 2021
https://www.linkedin.com/in/ankushmalik/-| -
ABOUT ME
EXPERIENCE
Hi, I am a software engineer. I love
to solve problems that affect people.
At the end of day, I want my work to
be my mark, to a certain extent.
Fear I have is that if I look back 5-6
years from now and see a career that’s —
mediocre.
I want my work to be something which
would make me really, really proud
of myself and feel like an impact. I love
to experiment and like to learn new
things. I am hardworking, team-oriented
and very adaptive.
ALIXIR | DEEP LEARNING ENGINEER
DEC 2020 - PRESENT | BENGALURU
• Research and develop production grade deep learning systems.
• Personally handled data acquisition operations at various client sites based
in India.
WORKERB | FRONTEND DEVELOPER
OCT 2019 - DEC 2020 | REMOTE
• worked on the core extension app.
• Personally handled client deals and user on boarding.
• Made various complex features like user action recorder/player.
ISRO | TRAINEE RESEARCHER, SPACE APPLICATION CENTER
EDUCATION
B.TECH IN COMPUTER SCIENCE
GURU GOBIND SINGH INDRAPRASTHA
UNIVERSITY- | New Delhi, India.
Cum. GPA: 7.5
LINKS
GitHub://AnkushMalik
LinkedIn://ankushmalik
upwork://malikankush
Twitter://iamankushmalik
LANGUAGES
English (Fluent)
Hindi (Mother tongue)
SKILLS
PROGRAMMING
MAR 2019 - JUL 2019 | Ahmedabad
• Neural Architecture Search for Semantic Segmentation.
• LSTM based NDVI prediction.
• N-shot Learning.
OPENSUSE | GOOGLE SUMMER OF CODE INTERN, HackWeek
MAR 2018 - AUG 2018 | Germany
• Improve people collaboration in the Hackweek tool
• Change post rendering algorithm such that every post has equal chances of
getting views from users.
• created Notification feature(backend + frontend) and mailing notifications to
relevant users.
PECUL | COFOUNDER
AUG 2017 - MAR 2018 | New Delhi
• Stand Alone created fullstack web application and deployed it to Amazon aws
EC2 instance.
RESEARCH & PROJECTS
Python • JavaScript • Ruby
BODH | PYTHON, PYTORCH, GDAL, JQUERY, FLASK
WEB
A smart web application which uses N(samples)-shot learning for
classifying without using Siamese Network.
A web application which is used to create ML model by a user without writing any
single line of code. Accuracy achieved around 96-99%.
HTML/CSS • JavaScript
Reactjs • Nextjs • Ruby on Rails
Flask • Node • Browser Ext
Ant/Material Design • React Hooks
DB used: pgsql, mysql, mongo, Firebase
AI
• openCV • Pytorch • numpy/pandas
• seaborn • sklearn • ANN/CNN • keras
Others
Git • Unity3D • Linux
NAS FOR SEMANTIC SEGMENTATION | PYTHON, PYTORCH, UNET,
DARTS(arXiv:-)
Finding best neural architecture for Semantic Segmentation of Satellite
Images
Real World problem solved: ISRO will be able save Millions of dollars every year by
automation of image segmentation process.
