Mô tả công việc
Tóm tắt công việc
The Data Team at Flodesk is part of our Engineering Team, based in our Da Nang office. We serve as the single source of truth for the company, providing data-driven insights that empower teams across Product, Marketing, Finance, and Engineering.
As a Data Analyst, you'll play a crucial role in translating raw data into meaningful insights, optimizing data pipelines, and maintaining a self-service BI ecosystem that enables teams to explore and understand data effectively.
What you will do
Data modeling and analysis: Own the dbt models and analysis behind Flodesk's business metrics, taking raw data to tested numbers that every report agrees on. Build automated checks that reconcile data across sources, and trace any gap down to the rows behind it. Most of your work will sit in the finance domain (MRR, retention, LTV/CAC, and revenue and cost reconciliation), so prior knowledge of SaaS metrics and subscription billing models will help.
Business partnership: Be the data partner for our business team, mostly the Finance Director and leadership. Turn their questions into well-defined metrics, help settle open definitions (say, how to count a member who converts and cancels in the same month), and build the decisions into the models.
AI-first analytics: Do most of your building and analysis through AI agents, and own the result: checking the logic, testing edge cases, and catching what the agent got wrong. You'll build the dashboards using AI agents and curate the metric definitions and metadata that keep MCP answers accurate.
Yêu cầu
2+ years as a data analyst or analytics engineer, working hands-on in SQL against a production data warehouse.
Advanced SQL and data modeling. You pick the right grain for each table and get snapshot and point-in-time logic right. Our stack is dbt on Redshift and ClickHouse, so familiarity with dbt and Python will help.
A fast learner with an open mind. Prior experience with subscription metrics such as MRR or LTV/CAC is welcome but not required.
Experience doing analytics work through AI agents (Claude Code, Cursor, or similar) is highly desirable, and building an analytics MCP server or semantic layer for LLM use even more so. Either way, you have the judgment to check an agent's output rather than trust it.
Rigor and healthy skepticism. When numbers don't tie, you keep digging until you can explain the gap, and you're upfront about caveats and definitional choices instead of presenting one tidy number.
Clear written and spoken English. You can explain a metric definition to Finance without jargon, and write documentation that people (and agents) can rely on.
Thông tin khác
Data Analysis
MS SQL
DBT
Python
AI (Artificial Intelligence)
AWS Redshift
ClickHouse
LLM
MCP
Claude
Cursor
Thông tin chung