Mohammad Ninad Mahmud Nobo
Full-Stack, AI & ML Engineer
Building scalable full-stack systems and real-world AI applications
I am a BUET CSE undergraduate with experience in full-stack products, backend systems, and applied AI. My current work includes ML-assisted software engineering, LLM-based web testing, and medical AI research.
About
Building systems that matter
I am a BUET CSE undergraduate focused on full-stack engineering, ML-assisted software development, and applied AI systems.
My research and project work focuses on LLM-based software testing, machine learning, and medical AI, with hands-on experience building end-to-end AI-driven systems and datasets.
I focus on end-to-end execution: problem framing, architecture, implementation, and measurable real-world utility.
What I bring
- I connect AI models to production systems, not isolated demos.
- I design for reliability in low-bandwidth and resource-limited contexts.
- I can move from a research concept to deployable product architecture.
Full-Stack Systems
Building scalable backend services and practical frontend experiences with production-ready architecture.
AI and Machine Learning
Developing intelligent systems with LLMs, deep learning, and practical ML pipelines.
Software Testing Automation
Designing LLM-assisted workflows for robust test generation and coverage improvement.
Medical AI
Building trustworthy AI systems for healthcare use cases with conflict-aware reasoning.
Education
School and College
Uttara High School and College
2008 - 2012
Rajuk Uttara Model College
2013 - 2021
Undergraduate
Bangladesh University of Engineering and Technology (BUET)
BSc in Computer Science and Engineering
Thesis:Web Testing Using Large Language Models
Coursework:DSA, Operating Systems, Computer Architecture, Database Systems, Software Engineering, Machine Learning, AI, Compiler Design
Research Interests
Projects
Selected Engineering Projects
View all on GitHubMindTrace
AI-powered dementia care platform delivering real-time support for patients and caregivers through an accessibility-first mobile experience.
Problem
Caregivers and patients need timely dementia support without depending on fragmented, manual coordination.
Solution
Built an end-to-end Android + Spring architecture with AI-assisted workflows, accessibility-first UX, and cloud-ready services.
Impact
Turned a complex healthcare support flow into a usable real-time system with demos covering both product experience and infrastructure.
- - Built end-to-end Android + backend architecture for real-time caregiver support.
- - Designed accessibility-first mobile UX tailored for elderly users.
- - Integrated AI-assisted assistance workflows backed by Spring services and cloud-ready infrastructure.
Architecture and Proof
Evidence
- - Mobile feature demo
- - Infrastructure demo
- - Public codebase
Artifacts
Feature Demo
Infrastructure Demo
Gemma VetCare
AI-assisted veterinary and livestock decision-support system engineered for reliability in low-connectivity field environments.
Problem
Rural livestock owners often have limited veterinary access and unstable internet, delaying treatment decisions.
Solution
Designed Android + backend decision support with resilient API behavior for intermittent networks and practical field workflows.
Impact
Demonstrated AI-assisted guidance that remains usable in low-connectivity conditions where conventional cloud-only tools fail.
- - Developed an AI-assisted Android workflow for veterinary and livestock decision support.
- - Implemented REST APIs optimized for low-connectivity and intermittent network conditions.
- - Structured backend data flow for practical field usage with robust request handling.
Architecture and Proof
Evidence
- - Feature demo video
- - Public repository
- - Field-oriented architecture
Artifacts
Feature Demo
SKILL HUB
Role-based coaching management platform with scalable REST APIs and robust multi-user access control.
Problem
Coaching workflows break down without clear role separation and reliable data operations.
Solution
Implemented a role-based platform and scalable REST API layer with SQL-backed multi-user flows.
Impact
Delivered a maintainable product foundation with clear boundaries between admin, instructor, and learner responsibilities.
- - Designed a role-based web platform with clear admin, instructor, and learner access boundaries.
- - Implemented scalable REST APIs and relational schema design for multi-role operations.
Architecture and Proof
Evidence
- - Role-based architecture
- - API-driven backend
- - Public repository
Artifacts
Remote Gardening System
Embedded automation system that monitors environmental signals and controls gardening operations through sensor-actuator workflows.
Problem
Manual plant monitoring is inconsistent and inefficient for remote or unattended garden setups.
Solution
Built a microcontroller-driven sensing and actuation loop to automate monitoring and control decisions.
Impact
Converted a manual task into a reproducible embedded automation workflow with hardware-level deployment evidence.
- - Developed an embedded automation loop for garden monitoring and control.
- - Integrated sensor-actuator logic on microcontroller hardware for real-world operation.
Architecture and Proof
Evidence
- - Hardware demo video
- - Embedded codebase
- - Sensor-actuator pipeline
Artifacts
Feature Demo
Research
Research and Thesis Work
I design and evaluate AI systems for software testing and medical decision support, with an emphasis on reliability, robustness, and practical impact.
Web Testing Using Large Language Models
Thesis-driven research focused on LLM-powered pipelines for generating, validating, and benchmarking functional web test cases.
Problem
Manual web test authoring is expensive and inconsistent, especially when converting natural-language requirements into robust functional tests.
Solution
Built an end-to-end LLM-based generation and validation pipeline with evaluation workflows for robustness and coverage.
Impact
Established a research-grade workflow that moves from natural-language intent to measurable test quality, not just generated scripts.
- - Designed and implemented an end-to-end LLM pipeline for automated web test generation.
- - Investigated test robustness and coverage behavior across diverse functional scenarios.
- - Built practical evaluation workflows for comparing generated tests against expected behaviors.
Tech Stack
Tags
Architecture and Proof
Evidence
- Curated supervised dataset
- Benchmarking framework
- Reproducible evaluation pipeline
Artifacts
- A curated supervised dataset mapping natural-language functional descriptions to test cases.
- An evaluation framework for quality, robustness, and coverage analysis.
MedCAR: Conflict-Aware Medical Reasoning
Medical AI research on conflict-aware multi-model reasoning for chest X-ray interpretation with trust-calibrated clinical decision support.
Problem
Medical AI predictions can conflict across models, creating unsafe uncertainty if systems cannot reason about disagreement.
Solution
Designed a conflict-aware reasoning layer with semantic reconciliation and confidence-calibrated abstention behavior.
Impact
Improves safety and trust by turning contradictory model outputs into auditable, confidence-aware clinical support recommendations.
- - Integrated heterogeneous AI components into a unified chest X-ray analysis pipeline.
- - Developed a conflict-resolution layer using semantic reasoning and confidence calibration.
- - Designed trust-aware abstention behavior to improve reliability under uncertainty.
Tech Stack
Tags
Architecture and Proof
Evidence
- Conflict-resolution workflow
- Confidence-calibrated decision layer
- Integrated multi-model pipeline
Artifacts
- A conflict-aware medical reasoning workflow for contradictory model outputs.
- A confidence-calibrated decision layer for safer AI-assisted clinical support.
Tech Stack
Technologies I work with
Programming Languages
Web Development
Android Development
AI / Machine Learning
Testing and Evaluation
Data and BI Tools
Databases
Systems and Tools
Contact
Let's build something together
I am open to internships, full-time opportunities, and research collaborations in full-stack engineering, applied AI, and intelligent systems.