Subbalakshmi T (Android ML Apps Lead)

• A professional with 11.4 yrs of experience in IT, including 8 yrs specializing in Android development and over 3+ yrs working with Data Science and Machine Learning. Proficient in Team Handling and Communication.
• Ability to complete all phrases of software development life cycle including Analysis, Design, Functionality development, Testing, Deployment and Support.
• Proficient in Object Oriented Design, Problem Solving, Complexity Analysis and Debugging.
• Experience in Data Analysis, visualization, statistics and Machine Learning Techniques, Good Knowledge on Tableau.
• Extensive Model building experience with Machine Learning algorithms by tuning the hyper parameters for a classification and Regression Problems.
• Knowledge on creating mature Data science pipelines that involves collecting, cleaning, transforming, analyzing, and deploying data to generate actionable insights
• Used Scikit-Learn and Tensor flow library to implement ML Models.
• Proficient Experience on Android SDK, MVVM, View Model, Life-Cycle Aware Components, Web Services, JSON, SQLite Database, UI Designing, Eclipse IDE, Android Studio.
• Able to Design the Applications from the requirement and known for Writing efficient Maintainable and reusable code that preserves privacy & security.
• Expert in the Interaction between Various Devices and different Versions of Android.
• Excellence in work ethics, self-motivated and quick learner.
• Expert in Parsing the JSON, XML, RESTFUL Services and Consumed the Web Service APIs by using AsyncTask, Work Manager, Mutable Live Data.
• Proficient in Integrated Social Media Networks like Facebook, Twitter, and Google+.
• Familiar in how to work with Google Maps, Push Notifications (GCM), Camera.
• Expert in publishing the apps into Playstore, generating Key store file with Signed APK.
• Expertise in Integration of Google Analytics, Play store referral and HasOffers.
• Worked on Docker Integration to run Android Test cases written by using UI Automator.
• Worked on Wearable Apps, Android with JNI and Android with ML Integration.
• Used Lottie View Animation Library for JSON Data Animations in Wearable Application.
• Proficient in Continuous Integration and Continuous Deployment.
• Good in Mentoring the Junior Developers.

To explore my expertise in software design and architecture to create highly scalable systems that meet customer needs by applying my technical expertise and leadership skills to lead a team of engineers in delivering high-quality technical solutions that dynamically works towards the growth of organization.

GitHub URL  https://github.com/LakshmiT/MachineLearning

• Twitter hate Tweets detection With the rapid rise of micro-blogging platforms like Twitter, Facebook, and Tumblr, communication among individuals has become more indirect and accessible. People from diverse educational backgrounds and cultures express their views on various aspects of life daily. This has led to conflicts between users, and as a consequence, the use of hate speech has emerged as a significant issue. Detecting such content manually on these platforms is an incredibly time-consuming task.
• Each tweet is divided into "hate speech" or "non-hate speech." This can be done manually or using an automatic pre-processing technique.
• preprocessing the text using Natural language Tool Kit(NLTK), Remove stop words, tokenizing the words. Feature extraction performing the lemmatization, trained a model using Logistic Regression and achieved an accuracy of 94%.

Done my own projects for used car prediction and fake job prediction using Machine Learning Models.

Zebra Device Manager - Zebra Technologies - April 2024 to July 2024

Technologies used: Android, Kotlin, Java, Jetpack, JUnit, espresso, Jacoco, Machine Learning

Work History: Involved in the Analysis and Development of the Quick Tiles feature for Zebra Device Manager using android Tile's framework, boosting customer engagement by 20%.

Implemented POC for building a model to predict the app performance (under different conditions such as load, battery usage, network speed, memory usage, etc.) by android Automation using ML.

Developed 80+ JUnit test cases, improving code accuracy by 40% in the Zebra Device Manager Application.

Involved and collaborated in daily stand-up meetings.

Stride (Scene Text Recognition in Device) - Samsung R&D Institute India-Bangalore(SRI-B) - Jan 2022 to 2024 Feb

Technologies used: Android 9.0 to Android 10.0, Java, Android Studio, TensorFlow Lite Models, ML Models, RNN, CNN, OCR Models, Android JNI, GitHub

Work History: Developed and integrated multiple ML models (text localization, OCR) which is supportable from S22 device model for enabling multilingual image text extraction for 6 languages. Implemented bounding box visualization for detected words, lines, and blocks. Evaluated model performance across 6 diverse language datasets using GitHub BatchTest, resulting in above 90% improvement in accuracy/speed. Contributed to the design of 2-3 key UI screens/elements. Resolved 30 critical bugs and 20 minor bugs identified by QA. Participated in 4-5 code reviews per day maximum. Coded in C++ for 4 modules related to image processing/ML inference. Collaborated effectively with the team.

GALAXY WEARABLE - Samsung R&D Institute India-Bangalore(SRI-B) - Nov 2019 to Jan 2021

Technologies used: Android 4.3 to latest version (Android 11.0), java, Kotlin, Android Studio, Gerrit, AIDL, Bluetooth Manager API

Playstore URL: https://play.google.com/store/apps/details?id=com.samsung.android.gearnplugin

Developed and maintained core features (Widget, Quick Settings, Labs, Find My Band) for Android-Galaxy wearable Bluetooth pairing.

Resolved 40 above HQ issues, updated 4 features/UI elements(Widget, Quick Settings in Advanced Features, Labs, Find My Band), created 35 above test cases using UATF and Docker, set up the Docker testing environment, and fixed 60 above HQ QA bugs.

ALIVE SHOT - Samsung R&D Institute India-Bangalore(SRI-B) - June 2019 to Nov 2019

Technologies used: Android 7.0 and above, Android JNI, Tessaract, OCR, Android Studio, P4V

Developed AAR files integrating .so libraries for Layout Detection and Script Detection projects, enabling image text identification (credit cards, business cards, etc.). Resolved 40 above integration bugs related to the Android JNI/C++ components.

PICTURE FRAME - Samsung R&D Institute India-Bangalore(SRI-B) - Feb 2019 to June 2019

Technologies used: Android 4.0 & Above, Java, Android Studio, P4V

Improved the Picture Frame app widget (available via the Galaxy Store) by resolving max 30 bugs reported by HQ QA, both within the app and in its integration with the Samsung Gallery. Participated in code refactoring and developed unit tests to enhance quality and maintainability.

STAY HEALTHY - Photon Interactive Pvt Ltd - May 2018 to Oct 2018

Technologies used: Android 5.0 to 8.0, Kotlin, Google Maps, Android Studio, Flurry SDK for Analytics

Play Store URL https://play.google.com/store/apps/details?id=com.legacyhealth.stayhealthy

Developed a location-based health and wellness app (targeting a defined service area via zip codes) with features like Feeling/Doing/Urgent Care Maps, condition filters, notifications, feedback, and social sharing. My contributions included creating 4 Widget Specification Documents, integrating Flurry SDK, participating in development and handled Maps, Notifications and Social Sharing. performing 3-4 code reviews per day, optimizing 4 modules, resolving Z QA bugs, and engaging in unit testing. Participated in team meetings about work status updates.

VALVOLINE INSTANT OIL CHANGE (VIOC) - Photon Interactive Pvt Ltd - Dec 2017 to May 2018

Technologies used: Android 5.0 to 7.0, Google Analytics, Crashlytics, Android Studio

Play Store URL https://play.google.com/store/apps/details?id=com.mobile.vioc

Provided maintenance support for the Valvoline Instant Oil Change (VIOC) app, which helps users find nearby locations, check wait times, access coupons, and manage their account. Responsibilities included implementing client-requested changes (analytics, auto-updating store locators, login validation), participating in testing, and distributing APKs to QA via Test Fairy.

BRAG IT - Ascendia - Oct 2015 to Sept 2016

Technologies used: Android 2.2 to 5.0, Android Studio, Web Services, Jira, Material Design,Jenkins, Gitblit

Playstore URL https://play.google.com/store/apps/details?id=org.ascendia.bragit&hl=en

Contributed to the development of BragIt, a social media app similar to Facebook, featuring event sharing, challenges, friend invitations, circles, and group discussions ("Gangups") based on interest groups. Responsibilities included analysis, designed 4-5 layouts, developed 4-5 features/modules (Registration, Login, Forgot Password, Brags/Challenges, Gangup, Interest Groups, User Profile, Settings, Chat), interacting with the API team, performed and written unit tests for main 5modules , fixed 30 above QA bugs, and participated in code reviews.

CARTRADE NCD - CarTrade.com - January 2015 to Oct 2015

Technologies used: Android 2.2 to 4.4, Eclipse, WebServices, SQLite, Activities, GoogleMaps

Playstore URL https://play.google.com/store/apps/details?id=com.cte.newcardealers&hl=en

Contributed to all phases of the CarTrade NCD (for CarTradeExchange dealers) project, which allows dealers to manage test drives and car sales information. Key responsibilities included participating in technical design sessions, delivering timely, high-quality code for 5 features/modules (New Car Leads, New Car Prices, Dealership Profile, Dealership Details, Dealer Interaction, Report Location, EMI Calculator), designing A UI screens, publishing the app to the Play Store, and performed 30-40 unit tests.

CARTRADE - CarTrade.com -

Technologies used: Android 2.2 to latest version, Eclipse, Web Services, SQLite, HasOffers, Google Analytics

Playstore URL

https://play.google.com/store/apps/details?id=com.cartrade.car&hl=en

Contributed to all phases of the CarTrade.com project, a leading Indian online platform for buying and selling new and used cars. Responsibilities included designing 8-9 UI screens, adhering to client standards and best practices, working under the development lead's guidance, and coordinating with the API team to integrate services. Key modules include New Cars, Used Cars, Car Prices, Sell My Car, My Listings, Used Car Shortlisting, and News.

Global Certifi cate in Data Science and AI
Completed course on Data Science and AI and Duration is 06/2021 to 03/2023.
Certifi cate
https://shorturl.at/KZRAO