Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Luca Cacciarru, a Software Engineer with a strong foundation in UX/UI design. Since 2021, I’ve been at VLK Studio delivering complex full-stack projects with a focus on code quality, TDD, and engineering best practices. In recent years I’ve specialized in AI Engineering: LLM integration in production, prompt engineering with structured outputs, generative UI with SSE streaming, agent orchestration with mastra AI, and AI evals—applied in products such as Foodlista.it, E.INS and PhotoSI. I bring a strong UX sensibility and a technical mindset focused on performance, scalability, and reliability of AI systems.…I'm Luca Cacciarru, a Software Engineer with a strong foundation in UX/UI design. Since 2021, I’ve been at VLK Studio delivering complex full-stack projects with a focus on code quality, TDD, and engineering best practices. In recent years I’ve specialized in AI Engineering: LLM integration in production, prompt engineering with structured outputs, generative UI with SSE streaming, agent orchestration with mastra AI, and AI evals—applied in products such as Foodlista.it, E.INS and PhotoSI. I bring a strong UX sensibility and a technical mindset focused on performance, scalability, and reliability of AI systems.

Luca Cacciarru

Full Stack Developer, Web Developer, Front-End Developer, +2





I'm Luca Cacciarru, a Software Engineer with a strong foundation in UX/UI design. Since 2021, I’ve been at VLK Studio delivering complex full-stack projects with a focus on code quality, TDD, and engineering best practices. In recent years I’ve specialized in AI Engineering: LLM integration in production, prompt engineering with structured outputs, generative UI with SSE streaming, agent orchestration with mastra AI, and AI evals—applied in products such as Foodlista.it, E.INS and PhotoSI. I bring a strong UX sensibility and a technical mindset focused on performance, scalability, and reliability of AI systems.…I'm Luca Cacciarru, a Software Engineer with a strong foundation in UX/UI design. Since 2021, I’ve been at VLK Studio delivering complex full-stack projects with a focus on code quality, TDD, and engineering best practices. In recent years I’ve specialized in AI Engineering: LLM integration in production, prompt engineering with structured outputs, generative UI with SSE streaming, agent orchestration with mastra AI, and AI evals—applied in products such as Foodlista.it, E.INS and PhotoSI. I bring a strong UX sensibility and a technical mindset focused on performance, scalability, and reliability of AI systems.

Available to hire

I’m Luca Cacciarru, a Software Engineer with a strong foundation in UX/UI design. Since 2021, I’ve been at VLK Studio delivering complex full-stack projects with a focus on code quality, TDD, and engineering best practices.

In recent years I’ve specialized in AI Engineering: LLM integration in production, prompt engineering with structured outputs, generative UI with SSE streaming, agent orchestration with mastra AI, and AI evals—applied in products such as Foodlista.it, E.INS and PhotoSI. I bring a strong UX sensibility and a technical mindset focused on performance, scalability, and reliability of AI systems.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

Italian

Fluent

English

Advanced

Work Experience

Full-stack developer and UX/UI designer at VLK Studio

January 1, 2021 - Present

Lead full-stack development and UX/UI design for multiple projects (Foodlista, E.INS, PhotoSI). Architect scalable front-end architectures (TypeScript, Vue, Nuxt, Tailwind) and robust back-end services (Node.js, Drizzle ORM, PostgreSQL). Implemented AI-powered features such as automatic translation with structured outputs, EU allergen detection with confidence scoring, and menu generation from images/PDFs/links. Designed human-in-the-loop workflows ensuring AI outputs are proposed but not saved without explicit user confirmation. Built agentic pipelines with mastra AI (conversational agents, multi-step workflows: scraping, extraction, validation, enrichment, save) with PostgreSQL memory and golden-set evaluation (~50 test cases). Emphasized dynamic model selection via Vercel AI SDK and robust error handling with schema validation and retries. Achieved multi-tenant isolation in prompts and context, and AI feature gating for Pro plans. Focused on performance, reliability, and scalable ar