# Gabriele Lanzafame

> Senior AI Engineer & Full-Stack Developer. Freelance, based hybrid in London, UK & Italy,
> working remotely across Europe. Builds generative AI, computer vision and agentic automation
> systems for advertising, media production and marketing agencies.

## About

Gabriele is a systems thinker who designs architectures and production tools for agencies,
letting them increase their presence, conversions and creative output using AI while cutting
costs versus traditional production. Work spans full-stack development, 3D, computer vision,
machine learning and generative AI - moving fast from exploration to production-grade tooling
without losing sight of the business goal.

Representative work includes an in-house inference platform for custom-trained diffusion and
LLM models - generating assets for social campaigns and TV ads, combining cloud computing, 3D
and generative AI, all orchestrated by a backend that spins up workers and serves a UI used
directly by designers. This kept production in-house: faster, cheaper and more reliable than
outsourcing to a traditional production company.

Another notable project: a voice AI agent built to waste scammers' time on the phone, keeping
them talking to a bot instead of a real victim.

Recognition: BIMA100 AI & Tech Pioneers (personal listing); the voice AI anti-scam agent project
has also won 4x Cannes Lions (1 Gold, 3 Silver, 1 Bronze), a Shorty Award, a Webby Award, and a
Gold at the British Arrows.

Gabriele holds a strong stance on using AI securely and ethically: data should be stored safely
and used only with permission, and is a firm supporter of open source - using open models and
tooling, and sharing open-source repositories from his own work.

## Services

- **Generative AI for asset production** - custom-trained diffusion and LLM pipelines producing
  on-brand assets for social campaigns and TV ads, with inference platforms and worker
  orchestration.
- **Computer vision for analysis & automation** - segmentation, detection and masking pipelines
  turning raw footage/imagery into usable signal; automated media QA and tagging; risk, PR and
  case analysis tooling.
- **Consultancy for MVPs & exploration** - technical feasibility and architecture reviews, rapid
  prototyping for new AI capabilities, strategy and analysis tools for decision-making.
- **Agentic & AI automation** - internal tools and automations, agentic workflows and
  orchestration, RAG and knowledge-grounded systems, cloud-native containerised deployment.

## Skillset

- Development: Golang, Python, TypeScript, FastAPI, Angular, Astro, Solid.js, Azure, AWS,
  Runpod, Docker, Kubernetes
- 3D: Blender, RealityScan, Litchfield Studio, 3D Gaussian Splatting, 3D object generation
- Computer Vision: Segmentation, Detection, Masking, OpenCV, Pillow, FFmpeg, Moviepy
- Machine Learning: CNNs, Vision Transformers, Random Forest, PyTorch, TensorFlow,
  Reinforcement Learning
- Generative AI: Transformers, Diffusers, LLMs, Diffusion Models, RAG, LangChain, LangSmith,
  Agentic Development
- Tools: ComfyUI, GitHub Actions, Jira, Float, Adobe Suite

## Who he works with

Startups, advertising agencies, media production companies, marketing teams, creative
independent labs, AdTech, and other free/early-stage creative teams - mostly small, fast-moving
teams that need to move before a bigger studio or agency could.

## Location

Hybrid - London, UK & Italy (by arrangement). Remote across Europe.

## Contact

- Email: hello.gabriele.lanzafame@gmail.com
- Book a call: https://calendly.com/gabriele-lanzafame/30min
- LinkedIn: https://linkedin.com/in/gabriele-lanzafame
- GitHub: https://github.com/gfratmct
- Website: https://gabrielelanzafamee.vercel.app

## Pages

- [Home](https://gabrielelanzafamee.vercel.app/): full landing page - about, services, skillset,
  industries, ethics stance, contact.
- [Privacy Policy](https://gabrielelanzafamee.vercel.app/privacy-policy)