Data Scientist Job Description Template

In plain language: Specifies the exact job title, seniority level, and which team or department the role sits within — establishing the scope of authority and career ladder placement.

Common mistake: Using a generic title like 'Data Scientist' without a seniority qualifier. Candidates cannot self-screen for fit, and the posting attracts both $80K and $180K salary expectations simultaneously.

In plain language: Lists the specific, measurable tasks the employee is expected to perform — model development, data pipeline ownership, stakeholder reporting, and research — distinguishing essential from ancillary duties.

Common mistake: Listing generic responsibilities like 'analyze data and build models' with no specifics. Vague duties prevent the job description from serving as a performance management baseline after hire.

In plain language: States the minimum education, years of experience, programming languages, and domain knowledge that are non-negotiable prerequisites for the role.

Common mistake: Requiring a master's degree or PhD for a mid-level role when those credentials are not genuinely necessary. This screens out qualified candidates and creates potential disparate-impact exposure under EEOC guidelines.

In plain language: Lists additional skills, certifications, or experience that would strengthen a candidate's profile but are not disqualifying if absent — giving recruiters a secondary filter.

Common mistake: Making the preferred qualifications list longer than the required list. When preferred criteria outnumber required ones, hiring managers effectively enforce unstated requirements, undermining the job description's legal defensibility.

In plain language: Identifies the direct manager, cross-functional stakeholders the role interacts with, and the team size and composition the employee will work within.

Common mistake: Omitting cross-functional reporting relationships entirely. A data scientist who unexpectedly reports to both an engineering lead and a product director will be misaligned on priorities unless it is documented upfront.

In plain language: States the base salary range or band, bonus eligibility, equity if applicable, key benefits, and whether the role is on-site, remote, or hybrid — including travel expectations.

Common mistake: Omitting the salary range where local law now mandates disclosure — Colorado, New York, California, and Washington state all require salary ranges in job postings. Noncompliance carries fines up to $10,000 per violation in New York City.

In plain language: Assigns to the employer all models, algorithms, code, data pipelines, and research produced by the employee in connection with their role — including work performed remotely or on personal devices.

Common mistake: Limiting IP assignment to work performed 'on company premises.' Data scientists frequently work remotely and build models on personal hardware — narrow language leaves IP ownership ambiguous and creates disputes at departure.

In plain language: Prohibits the employee from disclosing or misusing proprietary datasets, model architectures, customer data, or business intelligence during and after employment — and specifies data security obligations.

Common mistake: No explicit reference to data security policies or PII handling. For data scientists with access to customer datasets, omitting data-handling obligations creates compliance gaps under GDPR, CCPA, and HIPAA.

In plain language: Affirms the employer's commitment to non-discriminatory hiring and invites qualified applicants with disabilities to request accommodations — required under US federal law and equivalent statutes in other jurisdictions.

Common mistake: Copying an EEO statement from a generic template without updating the contact details. Candidates who attempt to request an accommodation and reach a dead email address create ADA liability exposure.

A complete data scientist job description covers role title and seniority level, reporting structure, six to ten core responsibilities with specific deliverables, separately listed required and preferred qualifications, compensation range and work location, IP assignment, confidentiality and data handling obligations, and an EEO statement. Missing any of these sections reduces candidate quality, creates compliance risk, or leaves gaps that become disputes after hire.

A job description is generally not a standalone contract, but it becomes legally significant when incorporated by reference into an employment contract as a schedule or exhibit. IP assignment, confidentiality, and listed duties in the job description can all be enforced when the employment contract explicitly references them. Courts have also used published job descriptions as evidence of employer representations in wrongful-termination and discrimination cases.

It depends on where the role is based or where applicants are located. Colorado, California, New York state, Washington state, and several municipalities now require salary ranges in job postings. For remote roles, you must typically disclose a range if any applicant could reasonably be located in a covered jurisdiction. Noncompliance penalties range from $500 to $10,000 per violation depending on the jurisdiction.

Python is the near-universal requirement for data science roles in 2025 and should appear in required qualifications for any level above entry. SQL is required for most roles that involve querying production databases. R is relevant for statistics-heavy or academic-adjacent roles. Spark or Scala matters for large-scale distributed workloads. List only languages the role will genuinely use in the first 90 days — inflated language requirements reduce the qualified candidate pool without improving hire quality.

A data scientist job description emphasizes exploratory analysis, model development, and deriving business insights from data. A machine learning engineer description focuses on deploying, scaling, and maintaining models in production systems — MLOps, API serving, and infrastructure. In larger organizations these are distinct roles; at smaller companies one person may cover both. Mixing the two into one posting without clear prioritization produces a role that attracts neither profile well.

No. A full-time employment job description establishes the terms of an employment relationship — with IP assignment, benefits, and at-will or notice-based termination. Using it for a contractor arrangement risks misclassification as employment, triggering back taxes, benefits liability, and penalties. Contractors should receive a separate Statement of Work or Independent Contractor Agreement that defines project scope, deliverables, and payment without establishing an employer-employee relationship.

Review and update the job description every time the role materially changes — new tools, new responsibilities, or a change in reporting structure. For active hiring, update it before each new posting cycle. A job description that is more than 18 months old is likely listing outdated tools (e.g., older ML frameworks) and may not reflect current compensation bands, creating mismatched expectations at offer stage.

Yes, for any data scientist accessing proprietary data or building models the company intends to commercialize. Without an IP assignment clause, the employee may retain rights to models, code, or algorithms they created, particularly for work done remotely or on personal devices. The IP clause in the job description should be reinforced by an identical or broader clause in the employment contract, executed before the first day of work.

In the US, federal contractors and employers with 15 or more employees must include EEO language covering race, color, religion, sex, national origin, age, disability, and veteran status. In the UK, the Equality Act 2010 covers nine protected characteristics. EU member states have varying requirements. At minimum, include a standard EEO statement with a working accommodation contact. For roles in multiple jurisdictions, use the most inclusive statement and localize as needed.