I am a Machine Learning Engineer with 5+ years of experience delivering scalable AI solutions across financial and large-scale enterprise environments at Meta and Citi Bank. I am skilled in Python, PyTorch, TensorFlow, and Transformers, with hands-on expertise in LLM fine-tuning (LoRA, PEFT), retrieval-augmented generation (RAG), and multimodal embedding systems. I have built ranking and recommendation models using transformer-based architectures (T5, DLRM++), developed fraud detection and credit risk ML systems under Basel III and IFRS 9, and led MLOps automation with Airflow, MLflow, and Kubernetes. I optimize real-time inference with Triton, ONNX, and TorchServe, work with Kafka, Spark, and PostgreSQL for data engineering, and emphasize interpretability and Responsible AI across production systems.

Vrushank Prasanna

I am a Machine Learning Engineer with 5+ years of experience delivering scalable AI solutions across financial and large-scale enterprise environments at Meta and Citi Bank. I am skilled in Python, PyTorch, TensorFlow, and Transformers, with hands-on expertise in LLM fine-tuning (LoRA, PEFT), retrieval-augmented generation (RAG), and multimodal embedding systems. I have built ranking and recommendation models using transformer-based architectures (T5, DLRM++), developed fraud detection and credit risk ML systems under Basel III and IFRS 9, and led MLOps automation with Airflow, MLflow, and Kubernetes. I optimize real-time inference with Triton, ONNX, and TorchServe, work with Kafka, Spark, and PostgreSQL for data engineering, and emphasize interpretability and Responsible AI across production systems.

Available to hire

I am a Machine Learning Engineer with 5+ years of experience delivering scalable AI solutions across financial and large-scale enterprise environments at Meta and Citi Bank. I am skilled in Python, PyTorch, TensorFlow, and Transformers, with hands-on expertise in LLM fine-tuning (LoRA, PEFT), retrieval-augmented generation (RAG), and multimodal embedding systems.

I have built ranking and recommendation models using transformer-based architectures (T5, DLRM++), developed fraud detection and credit risk ML systems under Basel III and IFRS 9, and led MLOps automation with Airflow, MLflow, and Kubernetes. I optimize real-time inference with Triton, ONNX, and TorchServe, work with Kafka, Spark, and PostgreSQL for data engineering, and emphasize interpretability and Responsible AI across production systems.

See more

Experience Level

Expert
Expert
Expert
Expert
Expert
Expert

Language

English
Fluent

Work Experience

AI/ML Engineer at Meta
September 1, 2024 - Present
Fine-tuned and deployed Llama 3–based models using PyTorch Lightning + FSDP + Deep Speed, optimizing distributed training efficiency across 1,000+ GPUs with a 35% reduction in memory overhead. Built multi-modal embedding pipelines and retrieval-augmented generation (RAG) workflows to improve content relevance and factual recall. Implemented evaluation automation and collaborated on on-device AI acceleration techniques.
Machine Learning Engineer at Citi Bank
January 1, 2020 - June 1, 2023
Developed ML models for fraud detection and credit risk scoring; built end-to-end data pipelines for 4TB+ daily data; implemented NLP for document parsing (KYC/credit); optimized inference via ONNX and Kubernetes; built IFRS 9 PD/LGD/EAD models with versioned registries; established real-time anomaly detection; integrated XAI dashboards for compliance.

Education

Master of Science in Computer Science at University of North Carolina at Charlotte
January 11, 2030 - December 19, 2025
Master of Science in Computer Science at University of North Carolina at Charlotte
January 11, 2030 - January 7, 2026

Qualifications

Add your qualifications or awards here.

Industry Experience

Financial Services, Software & Internet, Media & Entertainment, Professional Services