Valiotti Data is a team of data engineers and analysts working with fast-growing companies in SaaS, FinTech, and e-commerce: 7 years, 50+ delivered projects, clients with up to $2B in annual revenue. We build systems and hand the knowledge to the client's team instead of keeping it to ourselves.
What you'll do
Design warehouses and pipelines end to end, starting from the client's actual problem.
Make the architectural call and defend it, to the client and to the team, with the alternatives you rejected and why.
Set the engineering standard: code review, tech talks, conventions for SQL and Python, and how we configure Claude Code.
Build and maintain ETL/ELT pipelines with CI/CD for data workflows.
Run the technical side of client communication with the project manager: calls, written updates, scope.
Document models, pipelines and decisions so the client's team can operate everything after we leave.
Mentor mid-level engineers and unblock them on hard problems.
Must-haves
5+ years in data engineering, including at least one warehouse you owned from design to production.
Production-grade SQL and Python, the kind other people maintain.
Hands-on BigQuery, and Airflow 2.x or a comparable orchestrator.
Judgment on ClickHouse, Kafka, S3 and dbt: when to reach for them and when not to.
Production experience with an agentic coding tool.
English at B2 or above. You'll be on client calls from the first weeks.
You make architectural decisions without waiting for sign-off.
Nice-to-haves
Mentoring or code-review leadership for junior and mid engineers.
Public work: repos, case studies, talks.
A Claude Code or other LLM setup you've actually tuned, for example CLAUDE.md, slash commands, sub-agents.
How the work is structured
2 to 3 client projects in parallel. Each project team is small, usually one engineer and one analyst, so "lead" here means technical authority across projects and mentorship inside the company, not managing a squad on a single account.
You work with clients directly, and a project manager handles commercial questions and scheduling.
Fully remote, flexible hours. We work in the European time zone, with occasional calls with US clients.
What we offer
23 paid working days off per year.
Corporate Claude subscription.
English classes and reimbursed certifications.
Internal knowledge base: case studies, guides, architectural templates you don't have to build from zero.
Flat structure. Decisions take days.
How to apply
Optional, and it makes the first conversation much better: a link to code you're proud of, or a few lines about a Claude Code or other LLM pattern you use, a slash command, a prompt structure, a CLAUDE.md setup. It doesn't need to be original. Private repos are fine, we'll sign an NDA.
No test assignment. We read what you send, then one technical call and a final meeting with the founder.