Wolfgang Dafert
AI Prompting Evaluation
Experience
AI Prompting Evaluation
Powerfront
- Leveraged reasoning LLMs to design, optimize, and validate advanced prompt engineering strategies for complex use cases.
- Implemented Langfuse to establish robust prompt management and integrated AI evaluation pipelines for continuous performance monitoring.
Technical PM for GenAI
Pulsate
- Prototyped & validated concept: independently developed a fully self-coded PoC for an innovative marketing asset generator using OpenAI’s GPT Image 1, and set up Langfuse for evaluation.
- Led implementation & launch: directed the engineering team in building the production version and collaborated with the product head to define and execute the go-to-market strategy.
AI Consultant and Implementation
TSTPrep
- Designed and developed LLM-powered (GPT4o and Gemini) solutions for teacher support, automating complex student evaluations in a production solution.
- Implemented prompt engineering and evaluation frameworks to enable rapid iteration, resulting in a 50% reduction in content production time due to LLM usage.
- Developed n8n agent and agency swarm for internal business process automation.
AI Dev
Freelance
- Set up proof-of-concept pipelines to train and evaluate custom ML models for classification tasks, recognizing multiple ingredients in images.
- Gained experience with agentic GenAI projects in production environments.
- Deployed models to Vertex AI endpoints for testing, measured performance metrics, and iterated on improvements.
- Coordinated with cross-functional teams (designers, developers, marketing) to align AI features with user needs.
GenAI Consultant and Dev
Self-employed
- Use case consulting and corporate education (GenAI and Cursor IDE).
- Developed and deployed custom Retrieval-Augmented Generation (RAG) and chatbot solutions for clients utilizing fine-tuned LLMs hosted on Runpod and AWS SageMaker.
- Performed backend development in Python/FastAPI, integrating LangChain for streamlined query handling and Guardrails to enhance prompt security.
- Built open WebUI, Gradio, Telegram and several other AI chat frontends.
CTO
Unison Media
- Defined and implemented the technical vision and strategy for a New York–based media agency, overseeing the development and project management teams.
- Spearheaded the adoption of a Zoho intranet, SOPs, and training manuals to optimize staff efficiency, automate business processes, and streamline website development.
- Assisted the CEO and team with a successful migration to Zoho One.
CTO & Technical Lead
Sea.earth
- Led technical research and development for a blockchain-based (Algorand) environmental startup.
- Researched and delivered feasibility reports on utilizing smartphones for environmental data collection, resulting in a successful grant application for $50,000.
- Designed an architecture for storing blockchain data in the InterPlanetary File System (IPFS) to enhance data security and accessibility.
Project Manager
Cekaso
- Managed a team of 5+ developers to build custom POS software solutions for a leading European furniture purchasing group, resulting in a 15% reduction in order processing time.
- Played a key role in UI/UX design and client-facing project management, ensuring on-time and within-budget product delivery.
Summary
Self-organized and results-driven AI evaluation engineer & project manager with 11+ years of experience leading software development teams and 3+ years specializing in generative AI solutions. Proven expertise in LLM evaluation using LangSmith, Langfuse, and Weights & Biases, alongside strong capabilities in cloud platforms (AWS, GCP, Azure) and Agile project management. Skilled at shipping innovative AI applications and building robust evaluation pipelines to ensure quality, reliability, and measurable impact. Seeking a challenging remote role to deliver high-performing AI products and drive meaningful business results through technical depth and collaborative leadership.
Skills
Project Management
- Agile
- Tech Lead
- Cto
- Notion
- Clickup
- Jira
Generative Ai & Agent Frameworks
- Langchain
- Langsmith
- N8n Agents
- Langfuse
- Openai Products
Machine Learning & Engineering
- Evaluation
- Prompt Engineering
- Llm Fine-tuning
- Pytorch
Programming Languages
- Java
- Python
- Javascript
- Typescript
Cloud Technologies & Mlops
- Aws Sagemaker
- Ec2
- S3
- Gcp Vertex Ai
- Cloud Functions
- Azure Machine Learning Studio
Soft Skills
- Self Organizing
- Result Driven
- Remote Team Management
Languages
Education
TU Braunschweig
Master's Degree · Computer Science - Usability · Braunschweig, Germany
Similar Freelancers
Discover other experts with similar qualifications and experience