AI Engineer with less than a year in multimodal biomedical image analysis, MLOps, and LLM compressio
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
AI Engineer with experience in leading ML development for healthcare AI projects, focused on multimodal biomedical image analysis for glaucoma detection. Proven ability to develop ensemble ML models with high accuracy and F1-score, engineer scalable training/evaluation pipelines, and implement production-grade MLOps platforms using tools like AWS ECS Fargate and Docker. Skilled in low-bit precision techniques for LLM compression and building full-stack AI evaluation platforms.
Mohandas College of Engineering and Technology
B.Tech · Computer Science and Engineering (Honours in Machine Learning)
August 1, 2022 – June 30, 2026
Viswa Prakash Central School (CBSE)
Higher Secondary Education
June 1, 2020 – May 31, 2022
Indian Institute of Technology Kharagpur (IIT Kharagpur)
Research Intern
January 1, 2026 – March 1, 2026
Kharagpur, West Bengal, India
TelcoInsight: Enterprise Churn Prediction System & MLOps Pipeline
June 1, 2026 – Present
Engineered a production-grade MLOps platform for telecom churn prediction using XGBoost and Scikit-learn, achieving 82% Recall and 0.84 ROC-AUC. Developed asynchronous FastAPI REST APIs integrated with Pydantic validation and automated schema enforcement. Implemented enterprise-level data validation pipelines using Great Expectations to ensure feature integrity and inference reliability. Built multi-stage Docker containers automated through GitHub Actions CI/CD workflows and deployed workloads on AWS ECS Fargate.
BitNet: Low-Bit Precision Techniques for LLM Compression
June 1, 2026 – Present
Implemented BitNet 1.58 ternary quantization techniques for transformer-based Large Language Models using weights constrained to {-1, 0, +1}. Developed specialized CUDA kernels for low-bit matrix multiplication bypassing traditional FP16 computation paths. Achieved a 5x reduction in memory footprint and 4x faster inference speed on benchmark datasets. Integrated bitnet.cpp deployment pipelines demonstrating production-grade efficiency for compressed LLM inference.
AI Assistant Comparison Studio: Frontier vs Open Source
June 1, 2026 – Present
Built a full-stack AI evaluation platform comparing DeepSeek v4 Pro against self-hosted Llama 3.2 8B models deployed on AWS EC2 GPU infrastructure. Engineered a custom Retrieval-Augmented Generation (RAG) memory system using ChromaDB for persistent long-term conversational context. Developed a three-column security evaluation interface testing jailbreak resistance, system prompt hardening, and adversarial prompt mitigation. Implemented telemetry logging pipeline tracking latency, token usage, and evaluation metrics using SQL based analytics.
Machine Learning for Engineering and Science Applications
NPTEL
June 1, 2026 – Present
Reinforcement Learning
NPTEL
June 1, 2026 – Present
The candidate achieved a score of 68 out of 100, indicating a moderate level of proficiency with room for improvement in clarity, grammar, and professional language usage.
Limitations
Cultural Fit Analysis
The candidate's projects demonstrate a strong interest in cutting-edge AI technologies (LLM compression, RAG, MLOps) and a proactive approach to learning and applying advanced concepts. The diversity of projects, from enterprise churn prediction to biomedical image analysis and LLM evaluation, indicates adaptability and a broad technical curiosity. However, the lack of a psychometric score makes it impossible to assess cultural fit beyond technical alignment and project diversity. The candidate is still pursuing a B.Tech degree, which might indicate a junior profile despite advanced project work.
Soft Skills & Operational Fit
The psychometric test score of 0 is a significant red flag, indicating potential severe deficiencies in logical reasoning, work attitude, stress handling, and team collaboration. This raises serious concerns about the candidate's operational fit and ability to function effectively in a professional team environment. The English test score of 68 suggests moderate communication skills, which may require improvement for senior-level roles.
The candidate scored 0 out of 500, which is a critical failure. This indicates a complete lack of demonstrated proficiency in logical reasoning, work attitude, stress handling, and team collaboration, or an issue with test completion.
Limitations