Senior Data Engineer

Knak

About The Position

This is the first Data Engineer hire at Knak, and the role is foundational for Knak’s Data Infrastructure and AI enablement strategy. Snowflake holds the company's product usage, customer, pipeline, revenue, and behavioral data, but most of the company can't get to it without routing through one or two people. Our goal is that anyone at Knak can access the data they need from wherever they work through internal AI agents and champion-built tools. None of that is possible today because the data layer underneath doesn't exist. You will architect and own the governed query layer that fixes this: the per-department Snowflake views, the PII scrubbing logic, the schema documentation, and the connection standards that let analysts, AI agents, and downstream tools across the company query data safely and self-serve. This is a senior IC role with architectural ownership, not a maintenance role. The work is greenfield, cross-functional, and directly tied to a 90-day implementation plan with executive visibility. You will set technical direction in partnership with the Director of Data & Analytics, work directly with department leaders (VP Customer Success, CMO, VP Sales, CFO, Head of Security) to scope and prioritize view requirements, and partner with DevOps on connection patterns. You will be expected to make the hard architectural calls (what to expose, what to scrub, what to deprecate) and to defend those decisions to engineering and executive stakeholders. The technical foundation you build in your first six months will shape how every future agent, dashboard, and analytical product at Knak gets built, and you will help shape the data team as it grows.

Requirements

5+ years building production analytics systems on a cloud data warehouse (Snowflake strongly preferred; BigQuery, Redshift, or Databricks experience is transferable).
Deep SQL fluency. You can read a 200-line stored procedure and tell us what is wrong with it. You write CTEs and window functions without thinking. You have used INFORMATION_SCHEMA, EXECUTE IMMEDIATE, and exception handling in production.
Demonstrated architectural ownership of a data governance layer. You have designed role-based access, view-level scrubbing, or row-level security for a company at meaningful scale, and you can walk us through the tradeoffs you made and what you would do differently.
Production experience with a transformation framework (dbt, SQLMesh, or equivalent), version control discipline (Git, code review, CI for SQL), and a track record of introducing or significantly improving these practices on a team.
Working understanding of the BI and product analytics layer (Tableau, Looker, Mode, Mixpanel, or similar) and the judgment to anticipate how upstream warehouse decisions will affect downstream dashboards before they break.
Track record of mentoring and influencing peers. You have been the person other analysts and engineers go to for warehouse design questions, and you can point to specific examples of decisions you led that others adopted.
Strong written communication. The schema documentation, design docs, and decision records you produce are read by people who do not work with you every day, and by executives who need to understand the tradeoffs.

Nice To Haves

Experience with Fivetran or similar ELT tooling, especially AWS, MySQL, Salesforce, LinkedIn Ads, and product event sources.
Experience supporting AI agents or LLM applications as data consumers, including connection patterns, rate limits, and prompt-safe data shaping.
Python for data engineering (programmatic Tableau workbook edits, ad-hoc ETL, openpyxl, pandas).
Familiarity with PII handling under enterprise SaaS contracts.
Prior experience as the second or third data hire at a growing company.

Responsibilities

Power the AI Champion Program. The data layer you build directly determines whether the company's AI ambitions land or stall. Customer Success's automated EBR generation, Marketing's outbound qualification engine and lifecycle Slack bot, Sales' lookalike scoring, Finance's research agent, and Security's compliance matrix are all real projects, with real owners, currently blocked on governed Snowflake access. You will be the person who unblocks them, and the technical partner those champions work with as their tools mature from prototype to production.
Architect and own the governed Snowflake layer. Design, build, and maintain per-department read-only views in Snowflake covering product usage, customer accounts, pipeline, revenue, and behavioral signals. Make the architectural calls on materialization strategy, refresh cadence, and access boundaries. Each view is scoped to a department's use case, respects contractual restrictions on customer data, and excludes PII at the view layer.
Define connection standards for AI agents and applications. Partner with DevOps to publish connection patterns (authentication, rate limiting, error handling, approved deployment paths) so that AI agents and champion-built tools can consume governed data without ad-hoc Engineering tickets.
Encode PII scrubbing and compliance as code. Translate policy decisions on customer data, contract restrictions, and PII handling into view logic. View-level scrubbing is the primary control, ahead of any browser-level or tool-level redaction. You will work with Security to validate that views respect contractual constraints.
Document the warehouse. Publish schema documentation for every view: what fields exist, what they mean, how they connect to Salesforce, Tableau, Gong, and other sources, and what is excluded and why. This documentation is the contract between the data team and every downstream consumer.
Modernize the analytics codebase. Convert ad-hoc SQL, brittle stored procedures, and one-off pipelines into version-controlled, tested, modular models. Establish dbt (or equivalent) as the transformation layer. Introduce CI for SQL changes. Replace silent-failure stored procedures with observable, alerted pipelines.
Support the Tableau and Mixpanel layers. Several of our most-used dashboards (the Knak Product Analytics workbook, the AI OKR Tracker, the MAA/MAU rolling time-series) sit on top of Snowflake. You will own the data contracts those dashboards depend on, including the upstream extract pipelines, type-casting rules, and the daily snapshot architecture for behavioral metrics.
Set technical direction and raise the bar. Establish patterns, conventions, and standards that the next two to three data hires will inherit. Lead code review for SQL and transformation work. Mentor analysts across the company on warehouse design. Represent the data team in architectural discussions with Engineering and DevOps.