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

Hi there! I'm Kaleem Ullah, a software engineer and MSc student with 3+ years of experience in Python, machine learning, and full-stack development. I’m passionate about turning ideas into real-world AI systems, from data collection and modeling to polished front-end interfaces and deployable services. Right now I work with Fraunhofer IPA on Human Activity Recognition, Edge AI, and embedded ML for industrial applications. I enjoy solving practical problems, building end-to-end ML pipelines, and delivering robust technologies that people can actually use.…Hi there! I'm Kaleem Ullah, a software engineer and MSc student with 3+ years of experience in Python, machine learning, and full-stack development. I’m passionate about turning ideas into real-world AI systems, from data collection and modeling to polished front-end interfaces and deployable services. Right now I work with Fraunhofer IPA on Human Activity Recognition, Edge AI, and embedded ML for industrial applications. I enjoy solving practical problems, building end-to-end ML pipelines, and delivering robust technologies that people can actually use.

Kaleem Ullah

Full Stack Developer, Front-End Developer, Back-End Developer, +2





Hi there! I'm Kaleem Ullah, a software engineer and MSc student with 3+ years of experience in Python, machine learning, and full-stack development. I’m passionate about turning ideas into real-world AI systems, from data collection and modeling to polished front-end interfaces and deployable services. Right now I work with Fraunhofer IPA on Human Activity Recognition, Edge AI, and embedded ML for industrial applications. I enjoy solving practical problems, building end-to-end ML pipelines, and delivering robust technologies that people can actually use.…Hi there! I'm Kaleem Ullah, a software engineer and MSc student with 3+ years of experience in Python, machine learning, and full-stack development. I’m passionate about turning ideas into real-world AI systems, from data collection and modeling to polished front-end interfaces and deployable services. Right now I work with Fraunhofer IPA on Human Activity Recognition, Edge AI, and embedded ML for industrial applications. I enjoy solving practical problems, building end-to-end ML pipelines, and delivering robust technologies that people can actually use.

Available to hire

Hi there! I’m Kaleem Ullah, a software engineer and MSc student with 3+ years of experience in Python, machine learning, and full-stack development. I’m passionate about turning ideas into real-world AI systems, from data collection and modeling to polished front-end interfaces and deployable services.

Right now I work with Fraunhofer IPA on Human Activity Recognition, Edge AI, and embedded ML for industrial applications. I enjoy solving practical problems, building end-to-end ML pipelines, and delivering robust technologies that people can actually use.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Language

English

Fluent

German

Intermediate

Work Experience

Software Engineer at Third Gen Web

December 1, 2021 - January 1, 2023

Contributed to a modern e-commerce platform; built and maintained frontend components using Vue.js, effectively managing complex app states with Redux; optimized performance and reduced load times; coordinated UI delivery timelines with stakeholders.

Software Engineer (Studentische Hilfskraft) at Fraunhofer IPA

March 1, 2024 - Present

Remote work focusing on a Service Catalogue platform for the EU-funded AI-MATTERs initiative; coordinated API integrations, developed real-time dashboards using Angular and FastAPI, and translated Figma designs into responsive UI components. Implemented REST APIs with JWT authentication and wrote unit/integration tests.

Embedded Machine Learning Engineer (Studentische Hilfskraft) at Fraunhofer IPA

January 1, 2026 - Present

End-to-end ML pipeline for Human Activity Recognition; processed 10+ hours of multi-modal IMU and audio sensor data achieving 85% classification accuracy.