Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

Machine Learning Engineer and Data Scientist with 7+ years of experience building scalable AI/ML and Generative AI solutions across financial services and cloud environments. I specialize in LLM-powered applications, RAG pipelines, and agentic AI systems using LangChain, LangGraph, and CrewAI—delivering intelligent automation and decision-making at enterprise scale. I build end-to-end ML pipelines and deploy production-grade systems using MLOps, Docker, Kubernetes, AWS, and Azure, with strong experience in NLP, deep learning, predictive modeling, and compliance-focused engineering aligned to NIST/ISO/SOX requirements.…Machine Learning Engineer and Data Scientist with 7+ years of experience building scalable AI/ML and Generative AI solutions across financial services and cloud environments. I specialize in LLM-powered applications, RAG pipelines, and agentic AI systems using LangChain, LangGraph, and CrewAI—delivering intelligent automation and decision-making at enterprise scale. I build end-to-end ML pipelines and deploy production-grade systems using MLOps, Docker, Kubernetes, AWS, and Azure, with strong experience in NLP, deep learning, predictive modeling, and compliance-focused engineering aligned to NIST/ISO/SOX requirements.

Avinash Sonia

Data Scientist, AI Engineer, Data Engineer, +1





Machine Learning Engineer and Data Scientist with 7+ years of experience building scalable AI/ML and Generative AI solutions across financial services and cloud environments. I specialize in LLM-powered applications, RAG pipelines, and agentic AI systems using LangChain, LangGraph, and CrewAI—delivering intelligent automation and decision-making at enterprise scale. I build end-to-end ML pipelines and deploy production-grade systems using MLOps, Docker, Kubernetes, AWS, and Azure, with strong experience in NLP, deep learning, predictive modeling, and compliance-focused engineering aligned to NIST/ISO/SOX requirements.…Machine Learning Engineer and Data Scientist with 7+ years of experience building scalable AI/ML and Generative AI solutions across financial services and cloud environments. I specialize in LLM-powered applications, RAG pipelines, and agentic AI systems using LangChain, LangGraph, and CrewAI—delivering intelligent automation and decision-making at enterprise scale. I build end-to-end ML pipelines and deploy production-grade systems using MLOps, Docker, Kubernetes, AWS, and Azure, with strong experience in NLP, deep learning, predictive modeling, and compliance-focused engineering aligned to NIST/ISO/SOX requirements.

Available to hire

Machine Learning Engineer and Data Scientist with 7+ years of experience building scalable AI/ML and Generative AI solutions across financial services and cloud environments. I specialize in LLM-powered applications, RAG pipelines, and agentic AI systems using LangChain, LangGraph, and CrewAI—delivering intelligent automation and decision-making at enterprise scale.

I build end-to-end ML pipelines and deploy production-grade systems using MLOps, Docker, Kubernetes, AWS, and Azure, with strong experience in NLP, deep learning, predictive modeling, and compliance-focused engineering aligned to NIST/ISO/SOX requirements.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

Work Experience

Machine Learning Engineer / Data Scientist at Manulife

May 1, 2023 - June 1, 2026

Designed and deployed enterprise LLM/GenAI solutions including document intelligence, chatbots, summarization, and knowledge retrieval platforms to support large-scale automation. Built and scaled RAG architectures using embeddings and vector databases, improving retrieval efficiency by 40%+ and powering context-aware assistants for 300K+ users. Architected multi-agent AI systems (10+ agents) with LangChain/LangGraph/CrewAI for autonomous research/decision/validation workflows and orchestration. Developed predictive and generative models for risk intelligence, customer analytics, and text analytics using classical ML and deep learning. Implemented scalable end-to-end ML pipelines on TB-scale datasets using Spark and distributed systems. Applied MLOps with MLflow, Docker, and Kubernetes for training/versioning, monitoring, and continuous retraining. Deployed on AWS (SageMaker/EC2/S3/EMR) and exposed services via FastAPI/Django REST APIs, plus streaming pipelines using Kafka/Spark Stre

Machine Learning Engineer / Data Scientist at Scotiabank

January 1, 2021 - April 1, 2023

Built and deployed 5+ AI agents and early multi-agent workflows for customer support automation, fraud detection, document classification, and conversational AI. Developed ML/NLP/LLM solutions using transformer models and LangChain pipelines for sentiment analysis, fraud detection, document intelligence, chatbots, and semantic search. Created supervised/unsupervised predictive models (e.g., Random Forest, SVM, Logistic/Linear Regression, KNN, Naïve Bayes, XGBoost, K-Means, PCA, neural networks) improving fraud detection/customer targeting accuracy by 25%+. Built batch and real-time pipelines with Spark, Kafka, and Airflow for millions of transactions daily, including Kafka + Spark Streaming for near real-time insights. Deployed scalable services using Docker and Kubernetes (AKS) with FastAPI/Flask/Django REST APIs. Implemented MLOps with MLflow and CI/CD for model versioning, monitoring, logging, and retraining. Used AWS and Azure services (SageMaker, EC2, EMR, S3, Databricks, Synap

Data Analyst at Aveva

June 1, 2017 - March 1, 2019

Extracted and cleaned data using SQL and Pandas, handling inconsistent schemas and missing values. Performed exploratory data analysis (EDA) to identify patterns, anomalies, and trends, creating visualizations with Matplotlib and Seaborn for stakeholders. Built and automated data tasks (ETL and reporting) using Python scripts, SQL procedures, and Airflow, reducing manual effort. Performed statistical analyses, hypothesis testing, and KPI tracking to evaluate business performance and model effectiveness. Documented workflows and assumptions for reproducibility using Git for version control and collaboration.