Anda di halaman 1dari 3

DEBRAJ RAY

SARJAPURA, BANGALORE, KA 560037 | (C) 7829271722 | debrajpesit@gmail.com

https://www.linkedin.com/in/debraj-ray-8098b74a

Objective and Motivation


I want to work in an organization, which thrives on challenging problems, values differences in opinions,
and encourages the development of new ideas. Some of my positive attributes are perseverance,
dependability, and commitment to my responsibilities. I believe in the philosophy - If you don't have any
shadow, you are not standing in the light.

Skills
• JAVA, C++, Python • Apache Drill • Agile

• Android • Authentication – Kerberos, • Fast Learner


MapRSASL, Plain
• SQL • Innovator
• Encryption – SSL
• Linux • Team Leader
• Cloud computing – Apache
• Data structures STORM architecture • Team player

• Algorithms • Speech Recognition, CMU • Excellent communication


Sphinx, Nuance skills
• JDBC
• Machine Learning
• ODBC
• NLP

Work History and Contributions


Software Developer 02/2018 to
Magnitude Software Pvt. Ltd. – Bangalore Current
Team – Connectivity (previously, Simba Technologies)
https://magnitude.com/connectivity/
https://www.simba.com/

• The Drill ODBC driver of Magnitude is used by customers like Microsoft, MapR, Dremio, ODBC and
and Qlik, and I am responsible for its maintenance. Primary activities in this sphere include Apache Drill
– analysis of SalesForce tickets; fix or delegate issues to other developers; implement new
features as per client requirements; respond to clients in SalesForce; oversee the work of (Platform – C++)
other developers and ensure they are able to meet deadlines. I am the acting Team Lead
for the Drill Team at Bangalore.

• The driver release process at Magnitude requires extensive testing. A significant number of Automation
release tests are currently done manually. I am driving a project to automate most of the
release tasks which would save around four working days of two developers. This project
was started at my behest as the release tasks are quite monotonous and unimaginative for
the developers.
Learning at
• The learning curve at Magnitude is significantly steep. In a short span, I have learnt to set Magnitude
up Drill in a server, configure authentication protocols (such as plain-auth, Kerberos, and
MapR SASL auth), configure encryption (SSL, SASL encrypt), set up impersonation, connect
to drill-bit directly and via ZooKeeper, and many more.
Ramp-Up
• In the first two weeks at Magnitude, as part of a ramp-up project I developed a fully
functional JDBC driver, from an initial framework provided to me, which can process SQL
queries and fetch data stored in JSON format in a remote server. To ensure fast retrieval of
data, I implemented pagination with double buffering that reduced the fetch-time by 25%.

01/2015 to
Software Developer 02/2018
Motorola Mobility Pvt. Ltd. – Bangalore
Team – Moto Voice (Voice Assistant in Motorola Phones)
https://play.google.com/store/apps/details?id=com.motorola.audiomonitor&hl=en

• We have integrated Amazon's Alexa voice assistant with Motorola premium phones like Moto Alexa and
the Moto X series and the Moto Z series. I have contributed to the development of the Moto Voice
Moto Alexa architecture including the development of a speech authenticator module for
reducing false voice triggers and blocking imposters. I have also single-handedly migrated
the components of Moto Voice 2017 architecture for Alexa enrolment to Moto Voice 2016,
so that old phones like the first-generation Moto Z are compatible with the new Alexa
Mod.

• In the Lenovo Global Hackathon 2016, we conceptualized the integration of Moto Voice Moto Display and
into Moto Display (the ambient display on Motorola phones). This feature enabled Moto Voice
contextual voice commands on incoming notifications. For example, on a calendar
notification like "Dinner invitation," users can send an email reply to the inviter, start
navigation to the destination, call the inviter or book an Uber ride, directly from the lock
screen just by saying an available command. The prototype won the first prize in the
hackathon and is now in production. From conception to production, I have been the
principal developer of this feature.

• We revamped the architecture of Moto Voice 2016 and built Moto Voice 2017. The new Moto Voice
design loosely packed components that would work independently and in parallel on
streaming audio. Furthermore, any Motorola signature application will be able to
seamlessly connect to Moto Voice to enable speech recognition, voice authentication,
action recognition, etc. During the development of Voice 2017, I took ownership of
multiple components like ASR, TTS, audio logging, etc. I also created a wrapper interface
for clients to communicate with Moto Voice. The wrapper would ensure that each request
is tracked and responses are guaranteed to reach the right requester.

• I had the responsibility of maintaining and upgrading the grammar for accurate and natural Moto Voice
voice action recognition in different locales. I had also developed the Moto Voice 2016
enrolment UI. Besides, I fixed many stability issues like ANRs, app crashes, and tombstones,
which required a good understanding of the multi-threaded architecture of Moto Voice.
Education

Bachelor of Engineering: Computer Science


PES Institute of Technology (VTU) – Bangalore 2011 - 2015
• Graduated with 8.96 CGPA

PG Diploma on Machine Learning and AI June 2018 – June 2019


(Programming Language - Python)
IIIT – Bangalore
• Pursuing (part time)

Publications

• "Pipelining Acoustic Model Training for Speech Recognition Using Storm", published in the
IEEEXplore digital library, URL:
http://ieeexplore.ieee.org/xpl/abstractAuthors.jsp?tp=&arnumber=6663188
• "Sentiment analysis of mixed language employing Hindi-English code switching", accepted in
The International Conference on Machine Learning and Cybernetics (ICMLC) 2015, published in
IEEEXplore digital library, URL:
http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=7340934&url=http%3A%2F%2Fieeexplore.ie
ee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D7340934
• "Deviation from Trajectory Detection in Vision based Robotic Navigation using SURF and
Subsequent Restoration by Dynamic Auto Correction Algorithm", published in the MATEC
Web of Conferences, Volume 28, 2015, URL: http://www.matec-
conferences.org/articles/matecconf/abs/2015/09/matecconf_icame2015_04002/matecconf_icame20
15_04002.html
• "Immediate Listing of Business on Google Maps with Abstract Business Listing using a
Mobile Application", published by Enhanced Research Publications, URL:
http://www.erpublications.com/uploaded_files/download/download_25_11_2014_14_04_01.pdf

Patents

• Executing voice commands based on voice confidence and usage confidence metrics
Application number: 201731035636
(Patent is filed in India, and is yet to be published)

Certifications

• Oracle Certified Professional Java SE 6 Programmer


• Oracle Certified Professional Java EE 5 Web Component Developer

Anda mungkin juga menyukai