Cyprian Feliks - Senior AI/ML Engineer

Recommended expert

Copenhagen, Denmark

Experience

Jun 2023 - Dec 2025

2 years 7 months

Orlando, United States

Senior AI/ML Engineer

Andor Health

Led the design and development of the Digital Front Door system, a multi-agent voice AI assistant for patients, using Microsoft Agent Framework to orchestrate complex reasoning and conversational flows over LiveKit
Implemented a hybrid Retrieval-Augmented Generation (RAG) system using PgVector and HNSW-based indexing to handle user queries related to health and medical advice, reducing hallucinations and improving response relevance by approximately 20%
Built and deployed Model Context Protocol (MCP) tools such as Tavily Search tool and EHR retrieval tool to access real-time healthcare data, enhancing response accuracy by 90%
Pioneered the integration of HIPAA-compliant generative AI guardrails and a custom safety checker with TinyBERT into the production voice AI pipeline to ensure regulatory compliance and prevent toxic outputs
Developed an evaluation pipeline leveraging LLM-as-a-Judge and traditional metrics (ROUGE, BLEU, BERTScore) with Azure Evaluators, and established logging and monitoring using Prometheus and Grafana to provide real-time performance insights
Implemented MLOps and LLMOps pipelines on Azure Machine Learning with CI/CD via GitLab Actions and HITL (Human-In-The-Loop) using MLflow, achieving 99.9% uptime and scaling to millions of concurrent requests
Developed a multi-agent system using LangChain and LangGraph to orchestrate domain-specific fine-tuned LLMs for seamless cross-domain interactions and efficient information retrieval
Designed and executed distributed fine-tuning pipelines for open-source LLMs (Llama, Medical-Llama) using QLoRA and instruction tuning on Azure ML, accelerating convergence by 45% and reducing hallucination rates by 40%
Optimized inference performance through quantization, adapter fusion, and model distillation using ONNX, vLLM, and TensorRT, reducing response latency by 40% without compromising reasoning depth
Mentored and coached junior engineers by creating onboarding guides, leading knowledge-sharing sessions, and establishing best practices for generative AI, distributed systems, and healthcare-grade production readiness

Jan 2020 - May 2023

3 years 5 months

Chesterfield, United States

Senior AI Engineer

Amdocs

Collaborated with T-Mobile engineers and data scientists to develop AI solutions including a conversational AI chatbot, 3D reconstruction of antenna towers, and depth estimation models for physical metric measurement
Developed and deployed an AI-powered customer service system using fine-tuned GPT with QLoRA and conversational AI to improve customer engagement, resulting in a 30% reduction in response time
Optimized LLM performance and efficiency via model distillation, quantization, and pruning, achieving up to 30% latency reduction and faster inference for real-time interaction
Implemented a Retrieval-Augmented Generation (RAG) system with a FAISS-powered vector store to enhance customer query handling with more relevant and accurate responses
Optimized document management workflows on AWS S3 using LLM models for classification and summarization, reducing manual effort by 50% and increasing operational efficiency
Built end-to-end MLOps pipelines for real-time data ingestion on AWS SageMaker with MLflow for pre-processing and inference, improving antenna detection and monitoring speed by 40%
Engineered a depth estimation model using monocular techniques with MonoDepth and multi-scale fusion networks to enhance network optimization and planning precision
Applied GANs with digital twin technology to create 3D reconstructions from drone imagery of antenna towers, reducing modeling time by 30% and enabling real-time virtual inspections
Conducted exploratory data analysis and created Power BI visualizations to identify process bottlenecks, informing data-driven decisions and cross-department insights

Sep 2018 - Dec 2019

1 year 4 months

Copenhagen, Denmark

Hybrid

Machine Learning Engineer

SiteCore

Built a multivariate time-series forecasting system using a hybrid LSTM and Transformer architecture to predict commodity price movements, improving accuracy through attention-based models and feature engineering
Designed and implemented a K-means clustering model in Python with scikit-learn for customer segmentation, boosting targeted campaign conversion rates by 20%
Collaborated with data science teams to prototype and deploy recommendation systems using matrix factorization and autoencoders, optimizing algorithms for personalized user experiences
Created interactive Tableau dashboards surfacing KPIs such as model performance, conversion lift, and pipeline latency, empowering business teams for data-driven decision-making
Built and maintained ETL pipelines in Python on GCP using Google Cloud Storage for scalable data ingestion, transformation, and storage, ensuring reliable delivery and consistent recovery

Aug 2012 - Jul 2016

4 years

Copenhagen, Denmark

Software Engineer

SimCorp

Designed and implemented backend services using Django and Flask in a microservices architecture to ensure high performance and scalability for financial applications
Managed and optimized MongoDB and PostgreSQL databases for data integrity and reliability in large-scale transactions and financial reporting
Developed and maintained microservices using Java and Node.js, enabling seamless module interaction and smooth deployment versioning
Implemented containerization and orchestration with Docker and Kubernetes to improve development pipelines, reduce downtimes, and enable high availability
Built and maintained robust data pipelines using internal tools for log processing and reporting infrastructure, supporting company-wide analytics
Implemented A/B testing frameworks for front-end features using statistical methods to analyze user engagement and drive product decisions

Summary

Seasoned AI Engineer with over 10 years of experience, deep expertise in architecting and deploying large-scale AI systems across LLMs, multi-agent orchestration, computer vision, and cloud-native MLOps. Experienced in building production- grade pipelines using RAG, RLHF, distributed fine-tuning, vector databases, and automated CI/CD across Azure, AWS, and GCP. Demonstrated success in improving accuracy, reducing hallucinations, and accelerating deployment cycles through robust data pipelines, geospatial analytics, and model optimization frameworks. Skilled in translating complex research into scalable enterprise solutions while mentoring teams and driving cross-functional technical impact.

Skills

Programming & Tools – Python, Javascript, C++, React, Sql, Fastapi, Flask, Django, Restful Apis, Git, Jira
Machine Learning & Deep Learning – Supervised, Unsupervised & Self-supervised Learning, Cnn, Transformers, Tensorflow, Pytorch, Scikit-learn, Pandas, Pydantic, Spacy, Nltk, Transfer Learning, Multi-modal Models, Graph Neural Networks
Generative Ai & Llms – Gemini Pro, Gpt-4/3.5, Claude, Llama 3, Mistral, Command R+, Prompt Engineering & Versioning, Fine-tuning, Peft, Lora, Qlora, Instruction Tuning, Multi-agent Systems, Conversational Ai, Langchain, Langgraph, Crewai, Semantic Kernel, Microsoft Agent Framework, Tts & Stt Models
Llm Optimization – Vllm, Tensorrt-llm, Flashattention, Pagedattention, Dynamic Batching, Model Quantization, Peft (Lora, Qlora, Adapters), Distillation, Efficient Transformer Architectures
Rag & Vector Search – Hybrid Retrieval, Metadata Filtering, Pgvector, Pinecone, Faiss, Chroma, Hnsw Indexing, Multi-modal Rag, Context Re-ranking (Colbert, Minilm, Tinybert)
Mlops & Llmops – Aws (Sagemaker, Ecs, Eks, Lambda, Cloudwatch), Gcp (Vertex Ai, Bigquery, Cloud Run), Azure (Azure Ml, Azure Functions), Kubeflow, Mlflow, Llamaindex, Docker, Kubernetes, Helm, Terraform, Kserve, Triton, Model Monitoring & Ci/cd, Prometheus, Grafana
Data Science – Data Pipelines (Apache Airflow, Prefect), Spark, Postgresql, Mongodb, Nosql, Bigquery, Etl Pipelines, Eda, Power Bi, Databricks
Soft Skills – Technical Leadership & Mentorship, Cross-functional Collaboration, Strategic Thinking, Product-minded Engineering, Agile Team Collaboration, Continuous Learning & Research Adaptability

Languages

Danish

Native

English

Advanced

Education

Sep 2016 - Aug 2018

University of Copenhagen

Master of Science, Computer Science · Computer Science · Copenhagen, Denmark

Sep 2008 - Jul 2012

University of Copenhagen

Bachelor of Science, Computer Science · Computer Science · Copenhagen, Denmark

Need a freelancer? Find your match in seconds.

Try FRATCH GPT

Similar Freelancers

Discover other experts with similar qualifications and experience

Experience

Senior AI/ML Engineer

Andor Health

Senior AI Engineer

Amdocs

Machine Learning Engineer

SiteCore

Software Engineer

SimCorp

Summary

Skills

Languages

Education

University of Copenhagen

Master of Science, Computer Science · Computer Science · Copenhagen, Denmark

University of Copenhagen

Bachelor of Science, Computer Science · Computer Science · Copenhagen, Denmark

Similar Freelancers

Senior Software Engineer

Head of Risk Modeling & Cloud Infrastructure

Senior AI/ML Engineer

AI Agents Automation - LLM-Powered Agentic System

AI Engineer

Lead AI Scientist — Computer Vision & Sensor Fusion

Independent Contractor

Mentor

AR/VR/XR Architect

Cloud-Architect, Senior Solution Architect, Senior Software-Engineer

Workshop Leader 'Introduction to AI Development Tools'

AI – Data Specialist

AI Engineer

Product Owner AI Learning Platform

Team Lead – Engagement & Relevance

Freelance IT Consultant/Advisor

Principal (Data Scientist/Data Engineer/Gen AI Engineer)

AI Consultant & Advisor

Technical Lead Conversational AI

AI Tutor & Chemistry/Biology Specialist

MLOps Engineer

NLP/LLM Chatbot

Applied NLP: Word-Level Encoding for Smarter Event Predictions

Senior AI/ML Engineer

Data Scientist

AI Strategy Consultant

Senior Data Scientist

AI Engineer

External Service Provider

AI Solutions Project Lead Consultant