Hi, I'm Syed Zami-Ul-Haque Navid
A graduate student passionate about solving real-world problems
About
I completed my BSc in Computer Science and Engineering from Bangladesh University of Engineering and Technology. Since my graduation in February 2021, I had been working at Enosis Solutions as a Software Engineer. I worked there until June 2022.
Between January 2023 and July 2023, I was an SDET at Populate, a US-based health-tech startup.
I joined as a PhD student at Arizona State University in August 2023. Besides conducting research in Privacy of Android apps, I have been a GTA at
the School of Computing and Augmented Intelligence.
I have undertaken several research projects during my undergrad (and shortly after as well), mostly involving Artificial Intelligence.
As I have stated earlier, lately I have ventured into privacy and security of mobile applications.
Skillset
- Languages: C#, Python, C/C++, Java, VB.NET, working knowledge in R
- Scripting Languages: Bash, JavaScript
- Databases: MSSQL, Oracle SQL
- App Analysis Tools: Objection, Frida, MITMProxy, adb
- Machine Learning Frameworks and Libraries: PyTorch, Scikit-Learn, Numpy, Pandas, HuggingFace, OpenCV, OpenPose
- Development Frameworks: .NET, Angular, NodeJS
- Network Simulator: Cisco Packet Tracer, Wireshark, NS2
- Microcontroller Programming: Atmel Studio
- Ontology Tool: Protege
- Version Control: git, SourceTree, GitKraken
- Task Tracking: Jira, Confluence
I am also well versed in both written and verbal communication; be it with the team members or the clients. I am an active participant in idea generation sessions.
Experience
- I have taught CSE 110, An Introduction to Java Programming
- Performed Manual testing of both Development and Production environment
- Implemented Unit Testing
- Developed SDET handbook
- Technology: .NET
- Developed features for a California-based business management system
- Implemented functionalities on both front-end and back-end
- Wrote performant backend code following onion architecture
- Optimized Stored Procedures
- Designed and developed reports employing SSRS
- Technology: .NET, MSSQL, Angular, SSRS, SSDT, VB.NET
March 2022 - June 2022 | Software Engineer Level 2
Research
Investigating privacy breaching behaviors of Android apps
- We want to find out how instrusive each feature of an app is
- We have been analyzing network traffic
- We have also been collecting the API logs
- We also want to learn if apps behave differently in different geolocations
Malicious Code Detection Using Semantic Techniques
Fake News Detection from Bengali Facebook Posts about Covid-19
- We created a dataset containing Bengali Facebook posts regarding Covid-19 using CrowdTangle
- We trained Transformer based models on our dataset
- Having identified the best model, we applied that model on one year's worth (March, 2020 - June, 2021) of Facebook posts
- Presented analysis about the types and prevalence of Fake posts
- This paper was submitted to CHI 2022
Anomaly Detection from Videos
- We extracted two kinds of deep feature represtation from surveillance videos using I3D and OpenPose
- We proposed heirarchical Multiple Instance Learning to perform detection
- We emphasized on interpretability as well as accuracy of our system
- We used CRCV dataset for our purpose. We also intended to use their model for baseline comparison
Classification of Warnings Based on Source Code Features
- Our intension was to classify the warnings that static analysis tools provide after inspecting source code
- For features, we utilized several engineered properties of source code
- We applied state-of-the-art tree classifiers such as XGBoost and LightGBM
- Moreover, we used Decision Tree, SVM, Linear Regression and LSTM-based classifiers
Projects
An AI-based project that removes PII
An Itinerary Management App
- A webapp that helps tourists plan their itineraries
- This app would help tourists in choosing accommodation options as well as transports
- Every suggestion will be made keeping the tourist's budget in consideration
- This app was built with Java Server Faces (JSF) and Oracle RDMS
- Technology: JSF, Oracle RDMS
Clustering of similar music based on Collaborative filtering
- Given a spotify playlist, this system will recommend songs based on the perceived taste
- Similar songs were found by using K-Means algorithm and Eucledian distance
- The resultant clusters were illustrated using tSNE algorithm
- Moreover, the efficacy of the K-Means algorithm was determined with Silhouette method
Conversational AI trained on Facebook conversations
- To make a conversational bot, the OpenAI's GPT model was trained on a dataset of conversations
- The said dataset consists of Banglish conversations from Facebook Messenger
- Banglish is simply Bengali written with English Alphabet
- It was an experiment to see if the insights gained from pre-training on English can be transferred to Banglish
Education
Bangladesh University of Engineering and Technology
Dhaka, Bangladesh
Degree: Bachelor of Science in Computer Science and Engineering
- Operating Systems
- Computer Security
- High Performance Database Systems
- Computer Architecture
- Simulation and Modeling
- Fault Tolerant Systems
- Microcontroller and Microprocessors
Noteworthy Courseworks:
Tempe, Arizona
Degree: PhD in Computer Science
(Ongoing)
- Advanced Data and Information Privacy
- Android App Analysis
Noteworthy Courseworks:
Research:
Extra-curricular Certificates
Interests
- I am a beginner ukulele player. I suppose the guitar was way too big for me! So I wanted to start small. Literally!
- I would like people to believe that I can sing too! Watch me destroy two popular songs!
- I am also trying to learn how to play chess. It does seem daunting. Wish me luck!
- I hope to travel more