AT&T Data Science Interview Questions and Answers (2026 Guide)

Quick answer: Glassdoor's aggregate ratings for AT&T's Data Scientist (2.9 out of 5 difficulty, 60% positive) and Data Analyst (3.1 out of 5, 63% positive) roles describe a recruiter phone screen, a technical assessment or screening round, and a panel technical interview, though the exact round count varies by reviewer. AT&T also runs a genuine, large India Development Center across Hyderabad, Bengaluru, Chennai and Gurugram doing network-operations data analytics and AI/ML, but we could not find interview reports tied specifically to that India centre for Data Scientist or Data Analyst titles, only a general campus-style technical interview account.

About AT&T

AT&T is a US telecom carrier running wireless, broadband and the FirstNet public-safety network. Its India Development Center (IDC), founded in 2006, operates out of Hyderabad, Bengaluru, Chennai and Gurugram, growing from about 800 employees in 2023 to roughly 3,500 in 2025, with a stated target of 9,000 by 2027. (BusinessOfGCC) The centre's core work is network-operations software, BSS/OSS systems and cybersecurity, alongside a data-and-analytics function described as "AI analytics" supporting AT&T's US wireless, broadband and FirstNet operations, using Databricks, Snowflake, TensorFlow, PyTorch and Scikit-learn. (BusinessOfGCC) AT&T's own careers site lists dedicated Data Science and Data & Analytics job categories in Hyderabad and Bengaluru. (att.jobs/india) This is telecom-network-flavoured data work, closer to Jio's profile than to a consumer product-analytics team.

AT&T's Interview Process

The clearest sourcing we found is Glassdoor's aggregate data, though the underlying sample sizes are small and not India-labelled:

• Data Scientist track: rated 2.9 out of 5 for difficulty, 60% positive experience, averaging 11 days to hire, based on 8 submitted interviews. Reported stages start with a roughly 30-minute recruiter phone screen covering resume, background and motivation for applying, followed by a technical stage. Reviewers describe this technical stage differently: one account mentions a short 15-minute round centred on describing a past data science project and whether it was deployed to production; another describes a four-person panel video interview where one interviewer asks about past projects, a second asks for live Python coding, and a third asks technical questions on Python and PySpark. (Glassdoor; InterviewQuery)

• Data Analyst track: rated 3.1 out of 5 for difficulty, 63% positive experience. Reported process: a roughly 30-minute recruiter call, where the hiring manager sometimes joins to discuss the role and ask surface-level technical and behavioural questions, then a roughly one-hour technical screening (often over video call with screen sharing) on AT&T's data systems, ETL pipelines and SQL queries, plus a take-home assignment on product metrics, analytics and data visualisation, and questions on hypothesis testing, probability distributions and machine-learning fundamentals. Required skills listed: SQL, Python and Power BI. (Glassdoor; InterviewQuery)

What Candidates Report

The most detailed single account we could find is a GeeksforGeeks campus-drive write-up. It does not name the exact job title or confirm an India location, but the topics (numpy, pandas, model accuracy) suggest a data-adjacent technical role rather than pure software engineering. It describes two rounds: a technical assessment of 30 MCQs plus 5 coding questions in 90 minutes, open only to Computer Science and IT students, followed by a 60-minute interview over Microsoft Teams combining a technical segment (self-rating in Python and SQL, object-oriented programming questions on polymorphism and overloading versus overriding, AWS Lambda versus DynamoDB, writing runtime-polymorphism code, SQL queries, and coding challenges on counting duplicates and digit frequency) and an HR segment (walking through a past project, training and testing datasets, and libraries such as numpy and pandas). (GeeksforGeeks)

What We Could Not Verify

Glassdoor's own samples here are small, 8 submitted interviews for Data Scientist, and its Data Analyst sample size was not disclosed in what we could access, so the round-count variance between reviewers may reflect different teams rather than one fixed process. We found no interview account, on Glassdoor, AmbitionBox, Naukri Code360 or Blind, that ties specifically to AT&T's India Development Center for a Data Scientist or Data Analyst title; the GeeksforGeeks account above is the closest India-flavoured, data-adjacent report we found, but it does not confirm its job title or that it took place in Hyderabad, Bengaluru or Chennai rather than a different AT&T India entity. AmbitionBox did not return usable AT&T-specific results for either role. We also found no confirmed telecom-network-specific case study (for example churn or network-fault prediction) in any Data Scientist or Data Analyst interview account, even though the India centre's own described work is network-operations analytics, so we are not presenting that as a likely interview theme, only as a plausible one.

How to Prepare

• SQL and ETL fluency: multiple independent reviewer accounts, for both Data Scientist and Data Analyst tracks, mention SQL and data-pipeline questions directly.

• One deployed project you can describe end to end, including whether and how it reached production, since this came up as a specific round for the Data Scientist track.

• Core statistics for the analyst track: hypothesis testing and probability distributions are explicitly reported as assessed topics.

• Basic Python coding practice (string and array manipulation, OOP concepts) for any technical-round variant, and light AWS familiarity if the role sits closer to India's engineering-adjacent teams.