Beschreibung

I build intelligent AI systems powered by Generative AI, Deep Learning, and Machine Learning that solve real-world business problems and scale efficiently.

As an experienced AI/ML Engineer and Backend Developer, I specialize in creating LLM-powered applications, RAG systems, AI agents, Model building, and fine-tuning, along with the backend infrastructure required to deploy them in production.

From idea to deployment, I deliver end-to-end AI solutions that are accurate, scalable, and ready for real users.

What I Can Do for You

✔ Build Generative AI applications (LLMs, RAG, AI agents)

✔ Develop deep learning models for NLP, vision, and prediction tasks

✔ Create chatbots & conversational AI systems

✔ Design AI-powered automation workflows

✔ Develop high-performance backend APIs (FastAPI / Flask/ Django)

✔ Deploy and scale models using Docker, AWS, GCP

Generative AI

- LLMs (LLaMA, Transformers)

- Retrieval-Augmented Generation (RAG)

- Prompt engineering & fine-tuning

- AI agents (LangChain, LangGraph)

Deep Learning

- Neural networks (CNNs, RNNs, Transformers)

- Computer Vision & NLP

- Model training, tuning, and optimization

Backend Development

- REST APIs, microservices architecture

- Model serving & integration

- Scalable system design

Languages: Python, Java, C/C++

AI/ML: PyTorch, TensorFlow, Keras, Scikit-learn

Gen AI Tools: Hugging Face, LangChain, LangGraph

Backend: FastAPI, Flask, Django

Data: Pandas, NumPy, XGBoost, LightGBM

Streaming & Distributed Processing: Apache Kafka, Apache Spark, RabbitMQ

Databases: MySQL, PostgreSQL, MongoDB

Cloud & DevOps: Docker, Google Cloud Platform, AWS (ECR, EC2, S3)

Infrastructure & MLOps: MLflow, Apache Airflow, Jenkins

Monitoring & Observability: Prometheus, Grafana, LangSmith

Additional Experience

Mobile App Development: Native & cross-platform app development with AI/ML integration

backend integration (APIs, microservices, AI-powered features)

Sprachen

Englisch
Muttersprachlich oder zweisprachig
Deutsch
Grundkenntnisse

Arbeitsortpräferenzen

Vor Ort möglich

Munich (bis zu 50 km)

Julius Maximilians Universität Würzburg
Student Research Assistant
September 2023 - August 2024 (11 Monate)
Wurzburg, BY, Germany
Designed a multi-step agent architecture using LangGraph to decompose research queries, route to targeted sources (arXiv, local PDFs), retrieve relevant data, and generate structured summaries with inline citations.
Integrated automated evaluation (RAGAS, answer faithfulness, citation precision), achieving >80% factual consistency across 200+ benchmark queries.
Fine-tuned neural network model, balancing high predictive accuracy with efficient parameter adaptation
Containerized and deployed as a FastAPI service with async request handling, enabling researchers to process 10k+ documents and compress literature review cycles from days to hours.
Python, FastAPI, Pydantic, LLM, Agentic-AI, Gen-AI, RAG, LLaMA, Open-AI, Pandas, Pytorch, NumPy, Scikit-learn, Hugging Face, LangChain, LangGraph, LangSmith, AsyncIO, RESTful APIs, PostgreSQL.
Julius Maximilians Universität Würzburg
Applied AI / ML
DIGITALAGENTUREN & IT-CONSULTING
Mai 2025 - Dezember 2025 (7 Monate)
Wurzburg, BY, Germany
Implemented and benchmarked SOTA tabular representation methods (TabICL, SCARF, Class-Conditioned Contrastive Learning) against BERT and Autoencoder baselines across three IDS datasets.
Built end-to-end pipeline from preprocessing to embedding generation and training supervised/unsupervised models (RF, XGBoost, SVM, Isolation Forest, Neural Network), evaluated via AUC-ROC/PR and other metrics.
Assessed cross-dataset transferability and per-attack detection across DoS, Botnet, and Brute Force threats, delivering insights for identifying rare attacks in imbalanced network environments.
Python, NumPy, Scikit-learn, Machine Learning, Deep Learning, MLflow.
Python Deep Learning Machine learning NLP Pytorch
ThexSOl
Software Engineer
Dezember 2015 - Dezember 2021 (6 Jahre)
Islamabad, Islamabad Capital Territory, Pakistan
Built ML models for network threat detection and malware classification, including data preprocessing, feature engineering, training, and deployment.
Designed and developed an event-driven micro-services platform for real-time data processing with integrated ML inference for anomaly detection.
Implemented and deployed data collection agents to gather and analyze data, enabling insights for model development and system optimization.
Collaborated in cross-functional teams to redesign legacy architecture, replacing bottlenecks with resilient, horizontally-scalable services operated 24 × 7.
Python, Java, C, Django, Neural Network, Machine Learning, Microservies, Distributed Systems, Pytest, Sockets, RESTful APIs, Redist, Apache Kafka, PostgreSQL, Docker, Git, CI/CD, AWS.