Open to opportunities

Muhammad Irfan

@muhammadirfan11

I build production machine learning and LLM systems for real-world data, APIs, and deployment.

I've built production ML and LLM systems at Mobaix, including scalable pipelines for traffic-flow and incident data from ingestion through monitoring.

I developed an LLM-based natural-language data query system using Qwen and LangChain, with a RAG layer that explains database tables and supports controlled tool calls. I also containerized AI and REST API services with Docker for cloud and edge deployment.

At Fraunhofer ISE, I developed multimodal smart-building ML solutions for sensor failure detection, improving model performance by 11% through feature engineering, Random Forest, and Logistic Regression. I also modernized a legacy ML codebase and built database-backed REST workflows for sensor data.

Earlier at Sofy.AI, I delivered cloud-based ML evaluation pipelines and backend services in Python, Java, and Kotlin. I continue to build practical LLM applications, including RAG chatbots, patient pre-screening APIs, and computer-vision data augmentation projects.

Experience

Work history, roles, and key accomplishments

Machine Learning Engineer

Mobaix

Jun 2024 - Jun 2026 (2 years)

Designed and implemented scalable ML pipelines for data ingestion, preprocessing, model training, evaluation, deployment, and monitoring on real-world operational data. Built an LLM-based data query system and a RAG layer for natural language questions about traffic flow and incident data.

Machine Learning LLM RAG Docker REST APIs Data Pipelines Monitoring

Machine Learning Engineer

Fraunhofer Institute for Solar Energy Systems ISE

Apr 2020 - Jul 2023 (3 years 3 months)

Developed ML solutions for multimodal smart building data, combining numerical sensor measurements and textual information for sensor failure detection and monitoring. Applied feature engineering and statistical modelling to improve model performance by 11% using Random Forest and Logistic Regression.

Feature Engineering Statistical Modeling Random Forest Logistic Regression REST APIs Databases Python Machine Learning

Machine Learning Engineer

Sofy.AI

Jun 2017 - Sep 2019 (2 years 3 months)

Designed and implemented ML-based evaluation systems to generate application quality scores, enabling automated assessment and data-driven decision-making. Built and deployed cloud-based ML pipelines on Google Cloud for model iteration, evaluation, and hyperparameter optimization.

Machine Learning Python Java Kotlin REST APIs MongoDB MySQL