Help & Supportcareers@gravityer.com
AI Engineer with 1+ years in Machine Learning & AI Models Training
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Highly motivated Computer Science and Engineering student with a strong passion for Machine Learning and AI. Experienced in developing and deploying AI-powered solutions, including deep learning models for disease detection, multi-agent systems for content creation, and RAG-based chatbots. Proficient in Python, TensorFlow, PyTorch, and various AI/ML frameworks. Actively involved in co-curricular activities, demonstrating leadership and teamwork skills. Seeking opportunities to apply and expand expertise in AI engineering.
North South University
B.Sc. · Computer Science and Engineering (CSE)
August 1, 2020 – June 30, 2025
Enroute Digital
Digital Marketing Intern
January 1, 2025 – March 1, 2025
India
NSU ACM Student Chapter
General Member in Team Provision
May 1, 2022 – June 1, 2023
Dhaka, Dhaka Division, Bangladesh
IEEE NSU Student Branch
Core Volunteer
April 1, 2021 – June 1, 2022
Dhaka, Dhaka Division, Bangladesh
NLP-Based Sentiment Analysis on Product Reviews (IMDb)
June 18, 2026 – Present
Performed EDA on IMDb reviews, cleaned and tokenized text, and created three feature representations: TF-IDF, Word2Vec embeddings, and DistilBERT embeddings. Built and evaluated Logistic Regression models on all representations using accuracy, precision, recall, F1-score, confusion matrices, and ROC curves, achieving highest performance (88 percent accuracy) with TF-IDF.
View ProjectVoice Controlled Home Automation System with Arduino
June 18, 2026 – Present
Developed a smartphone app with voice recognition to receive commands such as "Turn on/off the lights." Implemented Arduino UNO to process Bluetooth-transmitted voice commands and determine the intended action. Integrated a relay module to control the power supply to the light bulb, activating it based on the interpreted commands.
Medicine Reminder System
June 18, 2026 – Present
Developed a mobile application using Android Studio and Java to function as an alarm for medicine reminders, doctor appointments, and medicine shopping. Utilized Node-MCU as the primary hardware component, integrated with an IR sensor to detect user presence for medicine tracking, and a buzzer for alarm notifications. Used Firebase for database management to track the number of medicines in stock.
Bonoful: Bengali Cinema Hall Website
June 18, 2026 – Present
Built a ticket booking platform for niche Bengali cinema using HTML, CSS, PHP, and MySQL. Designed backend functionality for user management and ticketing data.
Prison Management System
June 18, 2026 – Present
Developed Prison Management System (PMS) for efficient inmate, staff, and visitor management with role-based access control and secure record handling. Integrated features for tracking inmate movements, scheduling guard duties, managing visitor appointments, and using biometric authentication for enhanced security. Developed using PHP for backend, HTML/CSS for frontend, and MySQL for database management, aiming to improve facility operations, transparency, and security.
Iris Clustering
June 18, 2026 – Present
Applied K-Means, Hierarchical Clustering, and DBSCAN algorithms to segment the Iris dataset into distinct flower groups. Conducted detailed analysis of cluster validity using silhouette scores and dendrogram visualizations. Implemented data preprocessing and feature scaling to enhance clustering performance. Visualized cluster assignments and patterns through comprehensive plots for effective interpretation.
View ProjectPeople Flow Detection Using Object Tracking and Heatmap Visualization
June 18, 2026 – Present
Applied YOLOv11-seg for person detection and segmentation in video streams, integrating Norfair for object tracking. Captured movement coordinates to generate heatmaps highlighting high-density and frequently visited areas. Produced processed videos and visual reports for crowd flow analysis, congestion detection, and layout optimization.
View ProjectDiabetes Prediction with Machine Learning, Render Backend API and Streamlit Frontend
June 18, 2026 – Present
Evaluated multiple ML classifiers on the Pima Indians Diabetes Dataset and selected Decision Tree for deployment. Developed a FastAPI backend for predictions, containerized with Docker, and hosted on Render with automatic redeployments. Created and deployed a Streamlit-based frontend that connects to the backend API for real-time diabetes risk assessment.
View ProjectHeart Disease Prediction Using Machine Learning, FastAPI, and Docker
June 18, 2026 – Present
Trained a Logistic Regression model to predict heart disease risk based on metrics such as age, cholesterol levels, and blood pressure. Built a FastAPI backend for real-time predictions and containerized the application with Docker for scalable deployment. Deployed on Render with continuous integration from GitHub, enabling public API access via Swagger UI.
View ProjectParkinson's Disease Detection Using Deep Learning Method and Explainable AI
June 18, 2026 – Present
Utilized the Pre-VIT model to train MRI images for disease classification, accurately distinguishing between healthy individuals and Parkinson's patients, achieving nearly 98 percent accuracy. Applied the VGG19 model to train spiral hand-drawn images, classifying them into four levels of severity, and achieved an accuracy of nearly 97 percent. Applied LIME (Local Interpretable Model-agnostic Explanations) to interpret model predictions, with yellow highlighting the area responsible for the predicted result. Integrated the trained models into a Streamlit-based website for real-time image prediction and classification.
Fine-Tuning Transformers for QA Using BERT-Base-Uncased
June 18, 2026 – Present
Fine-tuned BERT-Base-Uncased on SQUAD for span-based question answering, implemented custom tokenization and answer mapping, and built an inference pipeline to accurately answer unseen questions while evaluating performance using Exact Match and F1 metrics.
View ProjectAI Agent for LinkedIn Caption Generation Using Prompt Engineering and OpenAI Model
June 18, 2026 – Present
Developed an AI agent that generates multi-language LinkedIn captions using LangChain's LLM-Chain and PromptTemplate, leveraging GPT-40-mini for structured and context-aware content generation. Applied advanced prompt engineering and temperature-tuned model parameters to achieve engaging, human-like captions, demonstrating practical LLM orchestration for professional content automation.
View ProjectA Multi-Tool AI Agent to Interact with Medical Datasets and Web Search
June 18, 2026 – Present
Built an AI agent that analyzes medical datasets (Heart Disease, Cancer, Diabetes) through custom SQL tools while retrieving real-time insights via SerpAPI-powered web search. Integrated LangGraph-based memory, multi-tool orchestration, and GPT-4.1 reasoning to create a contextual, data-aware medical assistant capable of structured analytics and knowledge-grounded responses.
View ProjectAI Agent for Article Scraping using N8N Workflow
June 18, 2026 – Present
Developed an AI-powered automation system that scrapes, summarizes, and emails article insights through a FastAPI backend integrated with an N8N workflow and Gemini Chat Model. Built a complete pipeline from user input to delivery, including Firecrawl-based content extraction, AI summarization, Google Sheet logging, and Gmail automation via low-code orchestration.
View ProjectMulti-Agent System using Crew AI for Instagram Content Creation
June 18, 2026 – Present
Developed an automated multi-agent system where research, content writing, reviewing, and image prompt generation agents collaboratively produced Instagram-ready captions and visuals. Implemented task-based coordination using CrewAI and LangChain for end-to-end content generation, enabling topic research, caption creation, quality review, and text-to-image prompt design.
View ProjectBangla RAG FAQ Chatbot
June 18, 2026 – Present
Built a Retrieval-Augmented Generation (RAG) based chatbot capable of understanding and answering Bangla queries using category-wise FAISS vector stores and SBERT embeddings. Integrated FastAPI backend with a Streamlit frontend for real-time text and voice interaction, enabling metadata-filtered retrieval and LLM-powered response generation.
View ProjectComparative Analysis of SOTA Deep Learning Models for Eye Cataract Type Segmentation and Classification
June 18, 2026 – Present
Utilized Roboflow to process and annotate a custom cataract dataset, classifying images into categories such as normal, nuclear, immature, and mature cataracts. Applied instance segmentation models (Yolo11, Yolov8, SAM2, Detectron 2) to identify cataract types and assess the proportion of the affected area in relation to the overall region. Developed a classification and segmentation pipeline, integrating advanced dataset preparation strategies to boost detection accuracy and overall performance.
AI Engineering Bootcamp for Programmers - Certificate of Completion
Ostad Online Academic Platform
November 1, 2025 – Present
Python (Basic) - HackerRank Certification Test Accomplishment
HackerRank
June 1, 2025 – Present
10th IEEE International WIE Conference on Electrical and Computer Engineering 2024 - Certificate of Acceptance Appreciation of Paper
IEEE
December 1, 2024 – Present
Cultural Fit Analysis
The candidate demonstrates a strong interest in AI and machine learning through a diverse portfolio of academic and personal projects, many of which align well with an AI Engineer role. The breadth of skills across different AI domains (NLP, CV, LLMs, agent systems) shows a versatile and curious mindset. While the experience is primarily academic, the initiative to build and deploy various AI applications suggests a proactive and self-driven individual. The volunteer experiences indicate a willingness to contribute to a community, which can be a positive cultural indicator. However, the lack of professional team-based project experience makes it challenging to fully assess cultural fit in a fast-paced, senior corporate environment.
Soft Skills & Operational Fit
The candidate's project descriptions indicate a proactive approach to learning and applying new technologies. Participation in student organizations suggests an ability to collaborate and contribute to team efforts. However, the professional experience is limited to internships and volunteer roles, which do not provide direct evidence of senior-level operational leadership or complex problem-solving in a corporate setting. The communication clarity in project descriptions is good, but the lack of professional experience makes it difficult to assess operational fit for a senior role.