Luka Margiani

AI and Automation Engineer specialized in designing production-ready AI systems and orchestrated workflows that solve real-world business problems. Expert in Python, n8n, and LLM APIs, with a focus on building RAG pipelines and autonomous agents. Proven track record of bridging the gap between AI and deterministic logic to ensure high reliability and operational efficiency in enterprise environments.

Frontend Developer & AI Automation Intern at Sifamo GmbH (2026-01 – 2026-04)

Architected complex HubSpot automation workflows and engineered digital infrastructure for AI summits while developing custom localization and email delivery solutions.

• Architected complex HubSpot automation workflows using conditional logic and custom registration forms, reducing manual ticketing and payment processing by 90%.
• Engineered and optimized digital infrastructure for AI summits, managing frontend and backend logic to maintain reliability during high-traffic events.
• Developed a custom localization engine for Webflow sites, utilizing JS/CSS to eliminate third-party plugin dependencies and improve site performance.
• Resolved technical bottlenecks in email delivery systems by auditing DNS, SPF, and DMARC records to maximize inbox placement for critical communications.
• Built and deployed high-performance Webflow sites, ensuring seamless integration with external hosting and third-party business tools.

AI automation & Consulting at TEPSA (Trans European Policy Studies Association) (2025-11 – 2025-12)

Deployed agentic pipelines for autonomous data extraction, executed large-scale ETL migrations, and engineered RAG-style tools for lead enrichment and content generation.

• Deployed five n8n agentic pipelines for autonomous data extraction and research using GPT-4-mini and Firecrawl, saving 10+ hours of manual labor weekly.
• Executed a large-scale ETL migration of 11,000+ records into a centralized Airtable CRM, significantly reducing data retrieval time and improving organizational observability.
• Engineered a RAG-style Spotlight Generator using Jina AI to scrape member news and draft social content, reducing manual drafting time by 80%.
• Built a GDPR-compliant expert research tool via Tavily and Dropcontact APIs, automating lead enrichment with 95%+ verification accuracy.
• Developed an n8n Error Logger that translates complex JSON payloads into plain English, improving system observability for non-technical stakeholders.
• Conducted a comprehensive digital audit to integrate AI tools like Opus Clip and Metricool, doubling content output without increasing headcount.

AI Automation & Web Development Intern at Sifamo GmbH (2025-04 – 2025-06)

Designed automated distribution workflows, implemented human-in-the-loop approval systems, and integrated AI tools for specialized image generation and marketing technology evaluation.

• Designed and deployed automated distribution workflows via n8n and Zapier, integrating OpenAI and LinkedIn APIs to reduce manual publishing time by 75%.
• Implemented webhook-based human-in-the-loop approval systems, ensuring 100% brand consistency and reliability for automated AI outputs.
• Integrated Hugging Face and Astria into production environments to automate specialized image generation, reducing reliance on external stock assets.
• Collaborated with the COO to audit internal systems and implement technical efficiency improvements based on deterministic logic and AI integration.
• Benchmarked 12+ emerging AI tools to develop a strategic implementation roadmap, reducing the evaluation cycle for new marketing technologies by 3 weeks.
• Developed custom WordPress solutions focused on responsive design and seamless user experience for executive-level digital presence.

Bachelor's Degree in Computer Science – SRH University of Applied Sciences in Berlin (2022-10 – 2026-04)