Nima Aminpour

Machine Learning Developer @AltaML

Calgary, AB, CA

Multi-Agent Workflow DevelopmentBuilt and deployed a full-stack AI-powered data analysis agent enabling users to query datasets in natural language via a chat interface. Developed specialized backend agents (planner, code generator/executor, and interactive clarifier) to handle complex queries. Designed an evaluation pipeline with LangSmith using LLM-as-a-judge metrics and synthetic data for workflow assessment- Retriever-Augmented Generation (RAG) Solutions: Deployed multiple RAG systems, including GraphRAG, with production-ready APIs and a Streamlit/React frontend. Designed an internal RAG package with reusable components and a full-stack app template, enhancing development efficiency and user feedback collection- Agentic Workflow Development: Engineered workflows for document ingestion and parsing using AutoGen, improving data extraction speed and accuracy for RAG implementations- Predictive Modeling for Financial and Operational Insights: Developed predictive models for financial indicators and customer spending behavior, boosting recall by 20%. Built a fleet maintenance model leveraging sensor and weather data, saving approximately $7M annually, and implemented healthcare workforce planning solutions with real-time insights in Power BI. Designed and deployed a time series model for now-casting economic indicators, allowing real-time insights into financial trends- Natural Language Processing: Created LLM and BERT-based models to classify documents by security tier, ensuring compliance, and developed a solution for maintenance work orders to identify equipment issues based on descriptions, streamlining operations- Computer Vision: Built object detection and segmentation models for identifying pavement distress, aiding proactive infrastructure maintenance.

Sharif University of Technology

Bachelor's degree, Engineering

University of Calgary

Master of Science - MS, Engineering