AI / ML / SYSTEMS / IDENTITY
Applied Machine Learning Systems Developer
Building practical ML systems from data to usable applications
I am a 3rd-year undergraduate focused on applied machine learning systems. My work sits between pure research and pure software engineering, with an emphasis on turning ML ideas into usable systems. Across my projects, I usually work through the full pipeline: data, feature engineering, model training, evaluation, APIs, and application-level integration.
Execution Pattern
System Focus Matrix
Data pipelines, feature generation, evaluation, and prediction APIs.
Biometrics, verification workflows, and monitoring systems.
Detection, counterspeech, and conversational outcome prediction.
Positioning
System-first
ML integrated with APIs, data flow, and usable applications.
Signal
Research + Build
Practical execution with enough research exposure to support the work.
AIML portfolio signal tuned toward systems, security, and deployable workflows
I am a B.Tech Computer Science and Information Technology student at UEM Kolkata, focused on applied machine learning and ML system design. I am not positioning myself as a pure researcher or a pure software engineer. My work is best described as applied ML systems building.
How I Build
Instead of treating machine learning as isolated experiments, I focus on end-to-end systems that connect: data -> features -> model -> evaluation -> API -> application.
Education
University of Engineering and Management (UEM), Kolkata
B.Tech - Computer Science and Information Technology
2023 - 2027
GPA: 8.7 / 10
Recognition
Amazon ML Summer School 2025
Selected participant in Amazon ML Summer School, reflecting strong interest and grounding in core machine learning concepts.
Main Technical Themes
Where the portfolio is strongest
A smaller but real research footprint centered on deepfake detection, multimodal analysis, and media authenticity.
Publication
IEEE INDICON 2025
IIIT Bangalore
PublishedD. Bhattacharya, S. Dutta, I. Mondal, D. Ghosh, A. K. Das, S. Bhattacharyya
This work studies multimodal deepfake detection by combining audio and visual analysis. The focus is on analyzing facial artifacts, speaker-related cues, and temporal inconsistencies within a scalable AI pipeline.
Forecasting services, verification pipelines, moderation systems, and media-authenticity workflows presented as applied ML system architecture.
Domain
System Node
A forecasting system for predicting product demand from historical sales data
Click to expand
Domain
System Node
A full-stack inventory and billing platform designed for small businesses
Click to expand
Domain
System Node
AI system to assist moderators in identifying harmful conversations and suggesting responses
Click to expand
Domain
System Node
A machine learning pipeline designed to detect manipulated audio and video content
Click to expand
Domain
System Node
An experimental biometric authentication system that verifies users based on voice characteristics
Click to expand
Domain
System Node
A concept platform designed to assist in verifying NGO claims and improving transparency
Click to expand
Domain
System Node
Explores how automated systems can detect impersonation or brand misuse across digital platforms
Click to expand
Current Focus
The strongest signal here is not raw project count. It is the repeated ability to structure ML work as systems with data flow, evaluation, APIs, and application integration. The next step is deeper proof through cleaner benchmarks and stronger deployment evidence.
Skills grouped around the way the work is actually built: models, APIs, systems, and implementation tooling.
Category
Category
Category
Category
Category
The path so far: education, system-building, one publication, and a shift toward more focused execution.
2023
UEM Kolkata
2024
Began building ML systems with backend integration, including forecasting, biometric, and verification-focused prototypes.
2025
Selected for Amazon ML Summer School.
2025
Research paper presented at IEEE INDICON.
2025
Built practical systems through hackathons and freelance ML/debugging work, gaining client-facing and implementation experience.
Present
Refining projects toward stronger implementation, better deployment evidence, and clearer technical positioning in applied ML systems.
Looking Ahead
The next version of the profile should be narrower and stronger: deeper applied ML systems, clearer deployment evidence, and better proof around forecasting, identity verification, AI security, and practical production work.
Open to ML systems discussions, collaboration, and practical technical work.
Primary Channel
Availability
Applied ML systems, research-adjacent collaboration, and backend-integrated ML work.
I am open to discussions about machine learning systems, research collaborations, and technical projects.