
| Job Type | Full-time, Permanent |
| Work Mode | Hybrid — structured, with fixed “anchor” onsite days set by the business |
| Seniority Level | Analyst-level, based on the JD’s own “Analyst Expectations” section — genuinely different from some other Barclays “Data Scientist” postings that sit at AVP level |
| Expected Salary | ₹12 – ₹22 LPA |
| Business Area | Decision Intelligence |
Data Scientist at Barclays, Gurugram sits within the Decision Intelligence team — worth knowing this specific team name, since Barclays runs multiple differently-scoped “Data Scientist” postings across different business areas (Controls, Risk, Decision Intelligence), and the actual work, seniority, and reporting line can vary meaningfully depending on which one you’re looking at.
What the Decision Intelligence team actually does
This team drives strategic business impact through advanced analytics, machine learning, and data science — developing predictive models, forecasting solutions, and analytical frameworks that shape customer, risk, and commercial decisions across Barclays. Data Scientists here translate business challenges into scalable analytical solutions, working closely with business, technology, and risk stakeholders to improve operational efficiency and strengthen risk management.
What the role actually involves
- Identifying, collecting, and extracting data from various internal and external sources
- Performing data cleaning, wrangling, and transformation to ensure data quality and analysis-readiness
- Developing and maintaining efficient data pipelines for automated data acquisition and processing
- Designing and conducting statistical and machine learning models to analyze patterns, trends, and relationships
- Developing and implementing predictive models to forecast outcomes and identify risks and opportunities
- Collaborating directly with business stakeholders to identify where data science can genuinely add value
Required experience
- Strong programming skills in Python and SQL
- Experience developing and deploying machine learning models, covering both supervised and unsupervised learning techniques
- Strong understanding of statistical analysis, experimentation, and predictive modeling
- Experience working with large, complex, and diverse datasets — including data preparation, feature engineering, and model optimization
- Familiarity with cloud-based analytics platforms, particularly AWS or equivalent environments
- Knowledge of software engineering and MLOps principles — version control, testing, and deployment practices
What genuinely strengthens an application
- Experience with Spark/PySpark and distributed data processing frameworks — called out as advantageous, not mandatory
- Ability to translate business challenges into analytical solutions and communicate technical concepts clearly to non-technical audiences
- Strong stakeholder management skills, with genuine confidence to constructively challenge and guide decision-making using data
- A commercial mindset genuinely focused on measurable business impact, not just technical elegance
- Real curiosity and a continuous learning mindset
What Barclays Gurugram actually offers
Barclays’ campuses provide modern workspaces, wellness rooms, on-site cafeterias, fitness centers, and collaborative meeting spaces — genuinely comprehensive facilities rather than a bare-bones office. The company also emphasizes structured hybrid working with clearly defined onsite “anchor days,” rather than fully remote or fully in-office arrangements.
A genuinely useful safety note from Barclays itself
As with Barclays’ other listings, the careers page includes a direct, prominent fraud warning — worth sharing with readers given how often large, recognizable banking names get used in recruitment scams targeting data and tech candidates specifically.
Questions Candidates Commonly Have
1. Is this the same seniority level as other “Data Scientist” postings at Barclays? Not necessarily — this posting’s own JD uses “Analyst Expectations,” while other Barclays Data Scientist listings (even in Gurugram) have used AVP-level expectations instead. Always check the specific expectations section in each posting rather than assuming seniority from the title alone.
2. Is remote work an option for this role? No — Barclays describes this as hybrid working with structured, fixed “anchor” onsite days determined by the business area, not fully remote.
3. Is Spark/PySpark experience mandatory to apply? No — it’s explicitly listed as advantageous, not a core requirement. Strong Python, SQL, and machine learning fundamentals are the baseline expectation.
4. Why do salary figures for “Data Scientist at Barclays” vary so dramatically across sources? Different platforms sample different populations — Levels.fyi’s data skews toward senior, more experienced submitters, while Glassdoor’s broader dataset likely better reflects the actual Analyst-level range this specific posting falls into. Confirm the actual offer directly with Barclays rather than anchoring to either extreme.
About Barclays
Barclays is a major British multinational universal bank, offering consumer banking, corporate and investment banking, wealth management, and credit card services across the UK, US, and numerous international markets. It competes with major global financial institutions including HSBC, Citigroup, Bank of America, and JP Morgan. Headquarters: London, United Kingdom Industry: Banking & Financial Services Ownership: Public company, dual-listed on LSE...
View Company Profile →Top Interview Questions
Prepare with commonly asked questions for this role
Supervised learning trains a model on labeled data to predict a known outcome — like predicting loan default risk from historical labeled data. Unsupervised learning finds patterns in unlabeled data — like clustering customers into segments based on behavior without predefined categories.
I'd start by understanding the business problem to identify which raw variables are likely meaningful, handle missing values and outliers systematically, create derived features that capture genuine signal (ratios, time-based aggregations), and validate feature importance empirically rather than including everything by default.
A data pipeline is built for repeated, automated execution — handling scheduling, error handling, and monitoring — while a one-off script is typically written for a single analysis without those production considerations. MLOps principles like version control and testing become essential once something moves from a script to a genuine pipeline.
Use a real, specific example — focusing on how you translated technical output into business impact and actionable next steps rather than technical detail, since this maps directly to the "communicate technical concepts to non-technical audiences" skill explicitly listed for this role.
I'd check for overfitting through proper train/test/validation splits, test performance on genuinely out-of-sample or recent data, consider model drift monitoring plans post-deployment, and ensure the model's predictions are interpretable enough for stakeholders to trust and act on.
