Help & Supportcareers@gravityer.com
AI Engineer with less than a year in AI/ML pipelines & RAG architectures
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Passionate Artificial Intelligence engineer from COMSATS University Islamabad, specializing in end-to-end AI/ML pipelines, Agentic AI systems, NLP, and RAG architectures. Demonstrated ability to deliver production-ready solutions through hands-on internship and academic project experience. Seeking to contribute technical expertise in a dynamic AI-driven environment.
COMSATS University Islamabad
Bachelor of Science · Artificial Intelligence
August 1, 2022 – August 1, 2026
EziTech Institute
INTERNSHIP
July 1, 2025 – September 1, 2025
India
Multi-Agent AI Shopping Assistant
June 1, 2026 – Present
Designed a multi-agent system with four specialized agents Scraper, Embedder, and Recommender agents coordinated by a central Orchestrator agent that detects product category, checks data availability, and triggers each agent in sequence. Integrated MongoDB for product storage, FAISS vector store for semantic similarity search, and Hugging Face models for embeddings and conversational recommendations.
AI Research Agent
June 1, 2026 – Present
Built an autonomous research assistant using LangChain and Qwen2.5-7B where a Search Agent retrieves real-time web data via Tavily API and a Critic Agent reviews and refines the final output for accuracy and quality. Used structured prompting for web content reading and report generation, producing well-formatted research reports with key findings, conclusions, and source URLs.
View ProjectMultimodal PDF RAG System
June 1, 2026 – Present
Built a PDF Q&A system that extracts text and images using PyMuPDF, converts them into unified embeddings using CLIP, and stores them in a FAISS index for semantic retrieval. Retrieves the most relevant text chunks and images based on user queries and passes them to Gemini 1.5 Flash, a multimodal model, to generate accurate context-grounded responses.
Skin Lesion Classification
June 1, 2026 – Present
Fine-tuned ResNet50 on the HAM10000 dataset to classify 7 types of skin lesions using Transfer Learning, freezing the base model weights and training a custom classification head on top. Applied data augmentation, class weighting, EarlyStopping, and learning rate scheduling to improve generalization and handle imbalanced data during training.
Credit Card Fraud Detection
June 1, 2026 – Present
Built a fraud detection system using XGBoost that analyzes transaction patterns including time of purchase, customer age, and distance between cardholder and merchant location to flag suspicious activity. Achieved near-perfect results with ROC-AUC of ~0.9995, Precision 0.927, and Recall 0.926, meaning the model correctly catches fraud while keeping false alarms low.
Multilingual Sentiment Analysis
July 1, 2025 – September 1, 2025
Built an NLP pipeline using transformer-based multilingual models (mBERT/XLM-R) for sentiment classification across multiple languages with text preprocessing and fine-tuning.
Custom AI Voice Cloner
July 1, 2025 – September 1, 2025
Developed a Text-to-Speech voice cloning system using Coqui TTS pretrained model, enabling custom speaker voice replication by extracting vocal embeddings and synthesizing natural-sounding speech from text input.
Ultimate RAG Bootcamp Using LangChain
Udemy
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's academic background in AI and a wide array of projects, including multi-agent systems, RAG, and deep learning applications, demonstrate a strong passion for AI and a proactive learning attitude. This aligns well with an innovative and research-driven AI engineering culture. The academic nature of many projects, however, suggests a need for more exposure to production-grade systems and team collaboration in a professional setting.
Soft Skills & Operational Fit
The candidate's internship description highlights a strong work ethic, punctuality, and consistent delivery of high-quality outputs, indicating good operational fit and reliability. The diverse project portfolio suggests an ability to work independently and manage complex tasks.