Help & Supportcareers@gravityer.com
Machine Learning Engineer with 1+ years in AI, NLP & Time-Series Forecasting
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 experience designing end-to-end ML pipelines for time-series forecasting, anomaly detection, and NLP. Proficient in Python, PyTorch, and Scikit-learn with strong expertise in feature engineering, model optimization, and handling imbalanced datasets. Improved prediction accuracy by 10% using hybrid LSTM-GCN models and achieved ROC-AUC 0.95 in anomaly detection systems. Pursuing M.Tech (AI & ML) at LNMIIT with focus on RAG for financial applications.
LNMIIT Jaipur
M.Tech · Computer Science (AI & ML)
August 1, 2024 – Present
JECRC University, Jaipur
B.Tech · Computer Science & Engineering
August 1, 2020 – June 30, 2024
Shodh AI (IndiaAI Initiative)
Machine Learning Research Intern
December 1, 2025 – June 30, 2026
India
LNMIIT Jaipur
Teaching Assistant
January 1, 2025 – May 31, 2025
India
Medical Chatbot RAG-based Q&A System
June 1, 2026 – June 1, 2026
Built an end-to-end Retrieval-Augmented Generation (RAG) chatbot over a 600+ page medical reference document, enabling natural language Q&A with context-grounded responses. Engineered a document pipeline using LangChain's text splitters to chunk PDF content (500-char chunks, 20-char overlap) and generated 384-dim embeddings via HuggingFace's all-MiniLM-L6-v2. Indexed embeddings into Pinecone's vector database for top-k similarity retrieval, integrated with Groq's Llama 3.3 70B LLM, and built a Flask-based web app with a real-time chat interface.
View ProjectCredit Card Fraud Detection System
January 1, 2026 – January 1, 2026
Designed an end-to-end fraud detection system on 284K+ transactions (0.173% fraud) using Isolation Forest, One-Class SVM, and LOF; achieved ROC-AUC 0.95. Refactored prototype into a modular, production-ready ML pipeline, improving scalability, maintainability, and reproducibility. Engineered 6 domain-specific features and tuned contamination parameters to match real fraud rates, improving precision-recall performance. Implemented a weighted ensemble model with normalized scoring and threshold tuning, detecting ~61% fraud cases (90/148) in unseen data. Deployed an interactive Streamlit dashboard for real-time prediction, batch processing, and analytics.
View ProjectHybrid LSTM-GCN Model for Stock Price Prediction
December 1, 2025 – December 1, 2025
Developed a hybrid deep learning pipeline combining LSTM and GCN on 6,500+ records for time-series and relational modeling. Constructed correlation-based stock graphs to capture inter-stock dependencies, reducing prediction error by 10% (MSE: 0.00144). Generated 20-day forecasts and evaluated performance using MSE and RMSE metrics.
View ProjectCultural Fit Analysis
The candidate's project diversity, ranging from financial forecasting to medical chatbots and material science, indicates a broad interest and adaptability, which aligns well with dynamic team environments. Their academic background in AI & ML, coupled with practical project experience, suggests a strong drive for continuous learning and application of cutting-edge technologies. The involvement in an IndiaAI Initiative internship further highlights a commitment to impactful work and innovation.
Soft Skills & Operational Fit
The candidate demonstrates strong problem-solving abilities, evidenced by their approach to complex ML challenges in projects. Their experience as a Teaching Assistant suggests good communication and mentoring skills, which are valuable for team collaboration and knowledge sharing. The refactoring of a prototype into a modular ML pipeline indicates an understanding of maintainability and reproducibility, crucial for operational fit in a production environment.