Someone leaves a large virtual warehouse running for a reporting tool that fires a query every few minutes, auto-suspend never kicks in, and the monthly Snowflake credit bill comes as a nasty surprise. Finding that, fixing it and stopping it from happening again is ordinary work for a Snowflake developer. The rest of the week goes on loading data, modelling it, writing SQL and Snowpark code, and setting up the roles and warehouses that decide who sees what and how much it costs.
Use this page to get a Snowflake developer job description you can adapt, along with what changes by level, how to test real platform skill, the common errors in these ads, and a view on contract and permanent hiring.
Snowflake developer job description template
Name your cloud (Snowflake runs on AWS, Azure and Google Cloud), your loading method and your transformation tool. If the job is a migration, say so in the first line, since those candidates search for that specifically.
About the role
[Company] is hiring a Snowflake Developer to build and run our Snowflake platform on [AWS / Azure / Google Cloud]. You will load data from [source systems], model it for [analytics / BI / data products] and keep queries quick and credit use under control. You will be based in [city / remote] and report to [data engineering manager or head of data].
Responsibilities
- Load batch and near real-time data with COPY INTO, stages and Snowpipe from [S3 / ADLS / Google Cloud Storage], or from Kafka through the Snowflake connector.
- Build incremental processing with streams and tasks or dynamic tables, choosing the right option per pipeline.
- Develop and test transformations in Snowflake SQL and [dbt], with Snowpark where [Python] logic is needed.
- Design schemas and data models for reporting, including dimensional models and semi-structured data in VARIANT columns.
- Size virtual warehouses, set auto-suspend and scaling policies, and separate workloads so loads and dashboards do not compete.
- Investigate slow queries with Query Profile, and apply clustering keys only where pruning justifies the cost.
- Track credit consumption by warehouse and team using ACCOUNT_USAGE views and resource monitors.
- Manage role-based access control, masking policies and secure data sharing in line with [company] governance rules.
- Maintain deployments through Git, CI and [Terraform or schemachange].
Requirements
- [X] years of data engineering or warehouse development, with hands-on production Snowflake work.
- Advanced SQL, including window functions, QUALIFY, MERGE and reading execution plans.
- Practical knowledge of micro-partitions, pruning, caching and how virtual warehouses consume credits.
- Experience with Snowpipe, streams and tasks in live pipelines.
- Experience with dbt or a comparable transformation approach, including tests and documentation.
- A clear grasp of Snowflake roles and privileges.
- Working knowledge of one major cloud and its storage and identity services.
Nice to have
- Experience migrating from [Teradata / Oracle / Hadoop / SQL Server] to Snowflake.
- Python with Snowpark, or Spark for processing that happens outside Snowflake.
- A SnowPro certification.
- Orchestration with [Airflow], and experience with data quality tooling.
What we offer
- [Salary range] and [benefits].
- [Work mode] and [on-call rota, if there is one].
- [Certification or training support].
How the Snowflake developer role changes by level
Junior Snowflake developer
Juniors write SQL, add sources to existing patterns and build dbt models under review. They should know warehouse basics and be careful with data, but they will not usually hold account-wide privileges. Strong SQL matters far more than platform years here.
Mid-level Snowflake developer
This developer owns pipelines and subject areas from raw load to final tables. They pick between streams with tasks and dynamic tables, handle schema changes from sources, tune slow queries and answer for the credits their work uses. Ask for examples they measured.
Senior Snowflake developer or expert
A senior designs the account itself: database and warehouse layout, role hierarchy, environment strategy, cost attribution and data sharing. They lead migrations and write the standards others follow. If the person will mainly design the platform across several technologies and rarely write code, you want a data architect instead, so title the post that way.
Snowflake developer skills that matter most, and how to check them
- SQL. Give them a table with duplicate events and late arrivals, and ask for a clean latest-record view plus a MERGE into a target. Note whether they reach for QUALIFY and whether they consider what a rerun does.
- Cost awareness. Ask why credits doubled last month and where they would look first. Good answers go to warehouse metering history, auto-suspend settings, idle warehouses, oversized warehouses and repeated dashboard queries. Candidates should also know that storage is charged separately from compute.
- Performance. Show a Query Profile with a full scan and an exploding join. Ask what they would change. A sharp candidate will tell you that scaling up helps a slow single query, while a multi-cluster warehouse helps concurrency, and that neither fixes bad SQL.
- Pipeline design. Ask them to design near real-time loading from files landing in cloud storage. Look for Snowpipe, a stream on the raw table, idempotent merges and alerts on failed tasks or stale streams.
- Security. Ask how they would give analysts access to customer tables with personal fields hidden. Expect functional and access roles, masking policies and an aversion to everyone using ACCOUNTADMIN.
A short exercise on a Snowflake trial account works well, since candidates work on the real platform and you can watch how they approach it.
Common Snowflake developer job description mistakes
- Treating any SQL developer as a Snowflake developer. Experience on Oracle or SQL Server transfers only partly, since indexes give way to micro-partitions and the cost model is entirely different. Fix: ask directly for hands-on Snowflake production work, or plan time to learn it.
- Bundling Snowflake and Databricks as equals. Candidates tend to be deep in one, and the daily work differs. Fix: name your platform, and list the second only if you truly run both.
- Calling the role a DBA. Snowflake handles most of the physical tuning and maintenance a database administrator would do elsewhere. Fix: if you want access control and cost management, write those duties down plainly.
- Leaving cost ownership unclear. Without a named owner, credit spend stays everyone’s and nobody’s problem. Fix: say whether this person answers for the bill, and what authority they have over warehouses.
- Listing every feature on the platform. Copying the product menu into requirements, from Cortex to Native Apps, puts off honest applicants. Fix: list what you use now and mark the rest as interest.
Contract, contract-to-hire or permanent: which suits a Snowflake developer
Contract developers suit migrations off Teradata, Oracle or Hadoop, a first account setup, or a focused cost clean-up. These have an end point, and someone who has done the same conversion before will avoid mistakes your team has yet to meet. Insist that role design and warehouse decisions are written down before they leave.
Contract-to-hire suits a company building its first data team. You get to see whether the person explains credit trade-offs clearly to finance and business users, which matters as much as technical depth.
Permanent hiring fits the longer work of running the platform as the business adds sources and consumers. If you need to hire Snowflake developers in India, Zilo AI fills the role on contract, contract-to-hire or permanent terms, shortlists by the exact services you run and screens for SQL, Spark and cost awareness. Our IT hiring overview covers data engineers and BI developers for the wider platform team.
Frequently asked questions
What should a Snowflake developer job description include?
State your cloud, how data arrives, your transformation tool and who the consumers are. Add whether the work is a build, a migration or platform upkeep, and who owns credit spend.
What is the difference between a Snowflake developer and a data engineer?
A data engineer works across pipelines and tools, often with Spark and orchestration, and may not use Snowflake at all. A Snowflake developer specialises in this one platform, including warehouses, roles, sharing and cost, although some overlap is normal.
What is the difference between Snowflake and Databricks roles?
Snowflake roles lean towards SQL, warehouses and managed data sharing. Databricks roles lean towards Spark, notebooks and Delta Lake, with more Python and machine learning work. Both platforms have moved towards each other, so check what your team actually does.
Does a Snowflake developer need certification?
It helps as a signal but should never replace a practical check. A SnowPro credential shows study, whereas a story about cutting credit use on a live account shows experience.
If you want a shortlist matched to your Snowflake setup, Zilo AI can do the search and first screening. Speak to our team about your cloud, your data sources and the project ahead.