Data Scientist and Data Engineer Resume Guide: Two Different Tracks Often Confused

Q: How do I tell if I'm "data scientist" or "data engineer" if my work spans both?

Pick the one that's more than 50% of your time. If you spend most time building models and analyses, you're DS even if you write some pipeline code. If you spend most time building data infrastructure, you're DE even if you do some analysis on top. Frame the resume to match the role you want; the bullets do the work of communicating which side dominates. If you want both kinds of jobs, maintain two resume versions.

Q: How important are Kaggle competitions for DS resumes?

Less than they used to be, but top placements still help. Grandmaster status or multiple top-10 finishes in major competitions add real signal. Mid-tier participation doesn't add much. For DS roles in 2026, business-impact bullets at work outweigh Kaggle placements; for new grads applying to first DS roles, Kaggle results can be primary signal in the absence of work experience.

Q: What about MLOps engineer or ML platform engineer?

Closer to ML engineer than to DS or DE. See our ML Engineer Resume Guide for that flavor. Distinguishing features: training-pipeline focus, model-serving infrastructure, MLflow/Kubeflow tooling, often more C++ than typical DE work.

Q: How do I frame switching from DS to DE or vice versa?

Lead with the bullets most relevant to the target role. Honest framing about why you're switching helps in interviews; the resume itself should emphasize work that translates. A DS targeting DE roles should highlight pipeline contributions, SQL depth, and ETL ownership rather than modeling work. A DE targeting DS roles should highlight any analytical projects, modeling collaborations, and statistical work — though this is harder if your DE work was purely infrastructure.

Q: Do I need a portfolio for DS / DE roles?

Helpful for new grads or career switchers; less important for experienced engineers. A clean GitHub with 1–2 substantive projects (a real analysis with public data, a pipeline that processes a non-trivial dataset) provides verification when work history is thin. Senior candidates rarely need a portfolio; the work history carries the resume.

Data Scientist and Data Engineer Resume Guide: Two Different Tracks, Often Confused

“Data scientist” and “data engineer” are two distinct career tracks that resumes commonly conflate. The work is different, the hiring bars are different, and what recruiters look for differs. Data scientists are evaluated on statistical rigor, modeling work, and business impact via analysis. Data engineers are evaluated on pipeline reliability, scale, and infrastructure work. This guide covers both tracks separately, what each resume needs to communicate, and how to position yourself if your work straddles the line.

Track 1: Data Scientist Resume

Data scientist roles split further into “analytical DS” (analysis-heavy, business-impact-driven, often closer to product analytics or growth) and “ML DS” (modeling-heavy, closer to applied ML engineer). Recruiters at FAANG, fintechs, consumer-product companies, and consulting hire substantial DS teams across both flavors.

What recruiters look for

Statistical rigor. Comfort with experimental design, A/B testing, causal inference, hypothesis testing.
Business impact. Bullets that tie analysis to decisions made, metrics moved, or strategy changes.
Tools. Python (Pandas, NumPy, scikit-learn), SQL fluency, sometimes R, often dbt or similar.
Communication. Bullets describing how findings were communicated to non-technical stakeholders.
Domain depth. Specific industry experience often matters (fintech analysts, growth-product analysts, marketing-mix modelers, etc.).

Strong DS bullets

Lead with business impact and methodological rigor:

“Designed and ran A/B test on the checkout-flow simplification (n=8M sessions, 4 weeks); identified 2.4% conversion lift; recommended rollout led to estimated $14M annualized revenue increase.”

“Built customer-LTV model (gradient boosted; n=180M customers) used by growth team for marketing-budget allocation; reduced CAC payback period from 14 months to 9 by enabling segment-targeted spend.”

“Led causal-inference investigation into apparent decline in Tier-2 user engagement; identified instrument-tracking bug responsible for 60% of the apparent drop; remaining 40% attributed to seasonality (validated via interrupted-time-series).”

What makes these strong: rigorous method, specific scale, clear business outcome, honest treatment of multiple causes.

Common DS resume pitfalls

Listing models trained without business outcome. “Built XGBoost classifier with 0.92 AUC” is a model you trained, not impact. What did the model do for the business?
Emphasizing dashboards as if they were modeling work. Building a Tableau dashboard isn’t the same as model work; frame appropriately.
Vague “data analysis” bullets. “Analyzed user behavior to identify trends.” What trends? What action followed?
Leading with notebooks rather than impact. Notebook-by-notebook descriptions of analyses don’t differentiate you. Lead with outcomes.
Padding with low-impact analyses. Quarterly reports that informed nothing don’t belong on the resume.

Sample DS resume bullet structure

EXPERIENCE
Stripe — Data Scientist, Risk Analytics                            2022 – Present
- Built quasi-experimental analysis framework for the risk team; identified 3 model deployments retrospectively responsible for false-positive rate increases that had been attributed to data drift
- Designed and ran A/B test on chargeback-prevention messaging (n=4M merchants); identified 22% reduction in chargebacks for treated cohort; recommendation now standard product flow
- Owned the customer-LTV modeling for the merchant-services team (gradient boosted model on 180M-row training set); used by sales team for prioritization
- Mentored 2 junior analysts on experimental design and causal inference

Track 2: Data Engineer Resume

Data engineer roles focus on building and maintaining data infrastructure: pipelines, warehouses, streaming systems, data quality. Closer to backend / infrastructure SWE than to data scientist work, despite the shared “data” naming.

What recruiters look for

Pipeline reliability and scale. Volume processed, throughput, freshness, uptime.
Tools. Spark, Flink, Kafka, Airflow, dbt, Snowflake, BigQuery, Databricks, sometimes Hadoop legacy.
Data modeling. Star-schema vs OBT, slowly changing dimensions, partitioning strategies.
Quality and governance. Data quality frameworks (Great Expectations, dbt tests), lineage tools, schema evolution.
Coding skills. Strong SQL, Python or Scala or Java for pipeline code.

Strong DE bullets

Lead with pipeline scale, reliability, and engineering improvements:

“Built event-streaming pipeline (Kafka, Flink, Iceberg) processing 280M events/sec across 3 regions; reduced ingest-to-query latency from 18 minutes to 2 minutes via materialized aggregations and per-partition checkpointing.”

“Migrated the company’s analytics warehouse from Redshift to Snowflake (140 TB, 800+ tables) over 9 months; cut query costs 47% and improved p99 query time from 11s to 2.4s.”

“Owned the data-quality framework for the marketing data org; reduced incident-causing schema breaks from ~6/month to <1/month via dbt tests, contract-based ingestion, and automated lineage tracking."

Common DE resume pitfalls

Treating it as DS-lite. DE work is engineering, not analysis. Don’t pad with analysis bullets if your job is pipeline work.
Tool-listing without scope. “Used Spark, Airflow, dbt, Snowflake” — what for? Specify the work and scale.
Missing reliability metrics. Pipelines that ran every day for 2 years without breaking is a real signal; bullets often miss this.
Vague “ETL” bullets. “Built ETL pipelines” is generic; describe specific pipelines, scales, and outcomes.

Sample DE resume bullet structure

EXPERIENCE
Snowflake — Data Engineer                                          2022 – Present
- Built event-streaming pipeline (Kafka, Flink, Iceberg) processing 280M events/sec across 3 regions
- Reduced ingest-to-query latency from 18 minutes to 2 minutes via materialized aggregations + per-partition checkpointing
- Owned the data-quality framework for marketing data org; cut incident-causing schema breaks from 6/month to <1/month
- Designed Slowly-Changing-Dimension Type-2 patterns now standard across 14 reporting tables

The Hybrid: Analytics Engineer / Full-Stack DE

Some roles blur the line — “analytics engineer” in particular sits between DS and DE. Resumes for these roles emphasize:

dbt mastery (most common framework)
Data modeling for analytical use (star schemas, OBT, semantic layers)
Stakeholder collaboration (working with DS and product teams)
SQL fluency at depth

If you’re an analytics engineer, frame as such — your resume reads neither as pure DS (no statistical rigor focus) nor pure DE (no streaming-systems focus).

Tech Stack Patterns

Data scientist (analytical)

“Python, SQL, Pandas, scikit-learn, R (some), dbt (some), Snowflake or BigQuery, Tableau or Looker, optionally Spark for large-scale work.”

Data scientist (ML-heavy)

“Python, SQL, PyTorch or TensorFlow, scikit-learn, Pandas, MLflow, often Spark, often Hugging Face for NLP.”

Data engineer

“Python or Scala, SQL (advanced), Spark, Flink or Kafka Streams, Airflow, dbt, Snowflake or BigQuery or Databricks, Iceberg or Delta Lake.”

Analytics engineer

“SQL (advanced), dbt, Snowflake / BigQuery / Redshift, Python (basic), Looker or similar BI, sometimes Airflow.”

Frequently Asked Questions

How do I tell if I’m “data scientist” or “data engineer” if my work spans both?

Pick the one that’s more than 50% of your time. If you spend most time building models and analyses, you’re DS even if you write some pipeline code. If you spend most time building data infrastructure, you’re DE even if you do some analysis on top. Frame the resume to match the role you want; the bullets do the work of communicating which side dominates. If you want both kinds of jobs, maintain two resume versions.

How important are Kaggle competitions for DS resumes?

Less than they used to be, but top placements still help. Grandmaster status or multiple top-10 finishes in major competitions add real signal. Mid-tier participation doesn’t add much. For DS roles in 2026, business-impact bullets at work outweigh Kaggle placements; for new grads applying to first DS roles, Kaggle results can be primary signal in the absence of work experience.

What about MLOps engineer or ML platform engineer?

Closer to ML engineer than to DS or DE. See our ML Engineer Resume Guide for that flavor. Distinguishing features: training-pipeline focus, model-serving infrastructure, MLflow/Kubeflow tooling, often more C++ than typical DE work.

How do I frame switching from DS to DE or vice versa?

Lead with the bullets most relevant to the target role. Honest framing about why you’re switching helps in interviews; the resume itself should emphasize work that translates. A DS targeting DE roles should highlight pipeline contributions, SQL depth, and ETL ownership rather than modeling work. A DE targeting DS roles should highlight any analytical projects, modeling collaborations, and statistical work — though this is harder if your DE work was purely infrastructure.

Do I need a portfolio for DS / DE roles?

Helpful for new grads or career switchers; less important for experienced engineers. A clean GitHub with 1–2 substantive projects (a real analysis with public data, a pipeline that processes a non-trivial dataset) provides verification when work history is thin. Senior candidates rarely need a portfolio; the work history carries the resume.