Sababa

Senior AI Engineer | NLP, LLM & Production ML Systems

Toptal • Freelance

Jan 2026 - Present • 4 mos

Worked with global clients through Toptal to design and deliver high-quality, production-ready AI systems across diverse domains. My work focused on building end-to-end machine learning and LLM-based solutions that go beyond prototypes—handling real-world challenges such as noisy data, ambiguity, and edge-case-heavy environments. I designed and implemented scalable pipelines covering the full lifecycle: data ingestion, preprocessing, feature engineering, model development, evaluation, and deployment. Key projects included NLP systems, document intelligence workflows, and LLM-powered applications using retrieval-augmented generation (RAG), structured prompting, and reasoning pipelines. A strong emphasis was placed on system reliability and performance. I built hybrid ML + LLM architectures that combine statistical models with LLM reasoning to improve robustness and accuracy in complex decision-making scenarios. These systems were evaluated using rigorous metrics (accuracy, macro-F1, etc.) and designed to maintain consistency under real-world constraints. In addition to modeling, I focused on clean architecture, modular design, and maintainability—ensuring that solutions are scalable and easy to extend. I collaborated closely with cross-functional teams, translating business requirements into technical solutions and delivering measurable impact through automation and data-driven insights.

AI Engineer & Data Scientist | NLP, ML & LLM

Upwork • Freelance

Jan 2020 - Present • 6 yrs 4 mos

Delivered 25+ successful projects with $50K+ in total earnings, building machine learning, NLP, and LLM-based solutions for clients across multiple industries. My work spans end-to-end AI system development, from raw data processing to deployment-ready applications. I have built solutions for document processing, predictive modeling, time-series analysis, and data analytics. Many projects involved working with messy, real-world datasets—handling missing data, inconsistencies, and class imbalance while maintaining strong model performance and reliability. A key area of expertise is developing LLM-powered systems, including RAG pipelines, document classification tools, and AI-driven automation workflows. I focus on designing systems that are not only accurate but also explainable and robust, using techniques such as SHAP-based interpretability and structured evaluation pipelines. Across projects, I have achieved strong results, including high accuracy on imbalanced datasets (~89%), improved forecasting performance, and efficient automation that reduced manual effort for clients. I prioritize clear communication, fast iteration, and delivering solutions that are practical, scalable, and aligned with business goals. Consistently maintained high client satisfaction by combining technical depth with a strong understanding of real-world problem-solving and production constraints.

Data Scientist

Remote • Freelance

Aug 2018 - Present • 7 yrs 9 mos

I am a data scientist and AI engineer with expertise in machine learning, deep learning, and natural language processing (NLP), currently pursuing a Master’s in Artificial Intelligence Engineering at Universität Passau. My work bridges academic research and real-world applications, with hands-on experience building scalable AI systems that solve complex problems. My recent projects include: * LLM & Transformer-based Tender Proposal Classification – Designed an end-to-end pipeline using HuggingFace Transformers, SHAP explainability, and Gradio deployment to automate tender categorization. * Automated Essay Scoring – Developed ML and LLM-based models achieving high scoring accuracy (Cohen Kappa 0.811) to streamline grading and expand access to education. * Passenger Volume Prediction – Built predictive models for bus ticket sales in Bavaria, improving operational planning for public transport. Previously, I worked as a Data Scientist & Software Development Manager at Proxima AI, where I combined statistical modeling, exploratory data analysis, and time series forecasting with leadership of software and data-driven projects. I bring strong proficiency in Python, R, BERT, HuggingFace, Scikit-learn, XGBoost, Tableau, and advanced ML frameworks, complemented by certifications in Data Science and Statistical Learning. I am passionate about transforming unstructured data into actionable insights, applying AI responsibly, and creating solutions that deliver measurable impact.I am a data scientist and AI engineer with expertise in machine learning, deep learning, and natural language processing (NLP), currently pursuing a Master’s in Artificial Intelligence Engineering at Universität Passau. My work bridges academic research and real-world applications, with hands-on experience building scalable AI systems that solve complex problems.