Help & Supportcareers@gravityer.com
AI Engineer with less than a year in Deep Learning, LLM Integration & Production API Development
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Machine Learning engineer with hands-on experience in deep learning, LLM integration, and production API development using PyTorch and FastAPI. Research background in federated learning, adversarial ML, and computer vision. Experienced building AI-powered backend services and deploying models for real-world applications. Strong mathematical foundation with national-level Physics Olympiad recognition.
Bangladesh University of Engineering and Technology
M.Sc. · Communication and Signal Processing
May 1, 2025 – Present
Bangladesh University of Engineering and Technology
B.Sc. · Communication and Signal Processing
November 1, 2019 – March 1, 2025
Bangladesh University of Engineering and Technology (BUET)
Graduate Teaching Assistant (EEE)
July 1, 2025 – December 1, 2025
Dhaka, Dhaka Division, Bangladesh
Bangla Document Q&A System (Multilingual RAG with OCR)
June 28, 2026 – Present
Built an end-to-end Retrieval-Augmented Generation (RAG) pipeline for Bangla and English PDFs, using multilingual sentence-transformer embeddings (paraphrase-multilingual-MiniLM-L12-v2) stored in a ChromaDB vector database. Added OCR support (Tesseract, Bengali language pack + Poppler) to extract text from scanned/image PDFs, with automatic detection of text-based vs scanned documents. Implemented hybrid retrieval combining semantic vector search with BM25 re-ranking, and generated grounded, source-cited answers via the Groq LLM API (Llama 3.1) in the question's language. Developed a FastAPI backend with a Streamlit frontend and containerized the full system with Docker (single image, offline-baked model) for reproducible deployment.
View ProjectElectronic Voting System using FPGA Board
June 28, 2026 – Present
Implemented a secure electronic voting system on an FPGA board, focusing on hardware design and Verilog programming.
View ProjectMovie Recommendation System
June 28, 2026 – Present
Built a collaborative filtering recommender using Matrix Factorization with SGD and L2 regularization on the MovieLens dataset to predict missing ratings
View ProjectSimulator of Damped Spring Mass System
June 28, 2026 – Present
Created a numerical simulator to model and analyze the dynamics of a damped spring-mass system using MATLAB.
View ProjectAI-Powered University Admission Assistant
June 28, 2026 – Present
Developed an AI-based admission assistant to help students determine suitable university programs based on their academic profile. Built a REST API using FastAPI for handling student queries and admission recommendations. Integrated LLM-based reasoning using the Groq API (Llama 3) to generate intelligent admission guidance. Designed backend architecture with Python, FastAPI, and SQLite relational database integration.
View ProjectUVC Disinfection Bot
June 28, 2026 – Present
Designed and prototyped a cost-effective autonomous UV-C disinfecting car using Arduino UNO, PID control, and sensors, with mapping and normal modes for navigation.
View ProjectFrequency Division Multiplexing Using Analog Electronic Components in PSpice
June 28, 2026 – Present
Simulated a frequency division multiplexing system in PSpice to study analog signal processing and communication techniques.
View ProjectM.Sc. Thesis Research
May 1, 2025 – Present
Developed a federated learning framework for multiclass DDoS attack detection under class-concentrated non-IID data, where standard FedAvg fails on minority attack classes (LDAP F1 = 0%). Proposed Class-Competence Aggregation (CCA), combining adaptive epoch scheduling and class-weighted loss; raised macro F1 from 62.05% to 91.46% and recovered the failing class from 0% to 86% F1. Built and trained hybrid CNN-LSTM and MLP models in PyTorch on the CIC-DDoS2019 dataset, validating results across 3 random seeds with ablation studies.
Undergraduate Research
January 1, 2023 – December 31, 2024
Proposed and simulated a Digital Twin-based framework for proactive handover in cellular networks, improving signal quality over traditional reactive methods
View ProjectCultural Fit Analysis
The candidate's diverse range of academic and personal projects, from federated learning to RAG systems and embedded systems, indicates a broad intellectual curiosity and a willingness to explore different technical domains. This adaptability and continuous learning mindset are positive indicators for cultural fit in a dynamic AI engineering environment. The focus on impactful projects like the Bangla Document Q&A System also suggests an interest in real-world applications and problem-solving.
Soft Skills & Operational Fit
The candidate's project descriptions indicate strong problem-solving abilities and a proactive approach to learning and applying new technologies. The academic research and personal projects demonstrate initiative and the capacity to work independently on complex technical challenges. The Graduate Teaching Assistant role suggests some experience in mentorship and coordination, which could translate to team collaboration skills, though direct evidence of operational fit in a corporate setting is limited.