SBI Data Scientist Recruitment 2026: 20 SCO Posts, Eligibility, Syllabus, Exam Pattern, Salary

Home › Blog › SBI Data Scientist Recruitment 2026

APPLY: 20 NOV TO 10 DEC 2026ADVT. CRPD/SCO/2026-27/31

SBI Data Scientist Recruitment 2026: 20 Specialist Cadre Officer Posts (Chief Manager, Manager, Deputy Manager)

State Bank of India has notified 20 regular Specialist Cadre Officer posts for Data Scientists working on GenAI, Agentic AI, RAG, MLOps, LLMOps, AI Security, AI Quality, Front end and Integration. This page covers the full notification in plain language: vacancy, eligibility, age, salary, selection process, the 60-question Professional Knowledge pattern, the complete syllabus, document specifications and a preparation plan.

20Total vacancies
10 Dec 2026Last date to apply
Rs 750Fee (Gen/EWS/OBC)
60 Qs / 100Professional Knowledge
70 : 30Written : Interview
Dec 26 / Jan 27Online test (tentative)
Apply on SBI Careers Start Preparing SBI Data Scientist Test Series & Notes

By Piyush Wairale (M.Tech, IIT Madras; former instructor, IIT Madras BS Degree Programme) · Published 21 November 2026 · Last updated 21 November 2026 · Source: SBI Advertisement No. CRPD/SCO/2026-27/31 dated 20.11.2026

Quick answer

SBI Data Scientist Recruitment 2026 (Advertisement No. CRPD/SCO/2026-27/31) offers 20 regular posts: Chief Manager (Data Scientist) 2 in SMGS-IV, Manager (Data Scientist) 14 in MMGS-III and Deputy Manager (Data Scientist) 4 in MMGS-II. Apply online at sbi.bank.in from 20.11.2026 to 10.12.2026; the fee is Rs 750 (nil for SC/ST/PwBD).

Eligibility needs a B.E./B.Tech/M.Tech in CS, IT, ECE, EEE, Data Science or AI & ML (or M.Sc Data Science/Statistics/CS, MA Statistics, M.Stat, MCA) with 60% marks, plus 3, 4 or 6 years of experience depending on the post. Age is reckoned on 31.10.2026.

Manager and Deputy Manager are selected through an online test (qualifying General Aptitude plus a 60-question, 100-mark AI/ML Professional Knowledge paper, no negative marking) and an interview, weighted 70:30. Chief Manager is selected through shortlisting and a 100-mark interview.

SBI Data Scientist Recruitment 2026: Overview at a Glance

OrganisationState Bank of India (SBI), Central Recruitment & Promotion Department, Corporate Centre, Mumbai
Advertisement numberCRPD/SCO/2026-27/31
Recruitment typeSpecialist Cadre Officers on Regular Basis
PostsChief Manager (Data Scientist), Manager (Data Scientist), Deputy Manager (Data Scientist)
Total vacancies20 (2 + 14 + 4)
GradesSMGS-IV, MMGS-III, MMGS-II
Application window20.11.2026 to 10.12.2026 (online only)
Eligibility cut-off date31.10.2026 (age, qualification, experience)
SelectionChief Manager: Shortlisting + Interview. Manager and Deputy Manager: Online Written Test + Interview (70:30)
Online testTentatively December 2026 / January 2027
Place of postingMumbai, or anywhere in India as per administrative requirement
Official websitesbi.bank.in/web/careers/current-openings

This is a specialist hiring drive for SBI's enterprise AI platform. The roles are built around production GenAI work: LLMs, Retrieval-Augmented Generation, agentic AI with tool calling and the Model Context Protocol (MCP), LLMOps, AI red teaming and testing of AI systems. If your experience is in classic analytics only, read the experience section carefully, because every post asks for hands-on AI/ML or GenAI work.

Important Dates for SBI SCO Data Scientist 2026

EventDate
Notification date20.11.2026
Online registration and fee payment start20.11.2026
Last date to apply and pay fee10.12.2026
Cut-off date for age, qualification and experience31.10.2026
Online written test (Manager and Deputy Manager)December 2026 / January 2027 (tentative)
Call letterTo be uploaded on SBI website; also sent by SMS and e-mail
InterviewTo be notified
Tip: registration is complete only when the fee is paid. You can save a partly filled form and edit it, but only three times, so keep all scans ready before you start.

The test may be just 4 to 8 weeks away (Dec 2026 / Jan 2027). Start now with the SBI Data Scientist Test Series & Notes made for this exact notification.

Start preparing now

SBI Data Scientist Vacancy 2026: Post-wise and Category-wise

Sl.PostNature / GradeSCSTOBCEWSURTotalPwBD (VI / HI)
1Chief Manager (Data Scientist)Regular / SMGS-IV000022VI 1
2Manager (Data Scientist)Regular / MMGS-III2111914-
Backlog-----VI 1, HI 1
3Deputy Manager (Data Scientist)Regular / MMGS-II000024-
Backlog10100VI 1
Total31211320Horizontal
  • PwBD reservation is horizontal and included within the parent category vacancy.
  • Vacancies are provisional and may change as per the Bank's requirement.
  • VI: Visually Impaired, HI: Hearing Impaired. Benchmark disability means at least 40% of a specified disability under the RPwD Act, 2016.
  • Reserved-category candidates for whom no reservation is shown may apply against Unreserved vacancies if they meet UR criteria.

Only 18 seats go through the written test (14 Manager + 4 Deputy Manager). With so few posts, the Professional Knowledge score decides everything. Practise with full-length SBI-pattern mocks.

Practise with SBI-pattern mocks

SBI Data Scientist Age Limit 2026 (as on 31.10.2026)

PostMinimum ageMaximum age (UR)Born between (UR, approx.)
Chief Manager (Data Scientist)27 years37 years01.11.1988 to 31.10.1999
Manager (Data Scientist)26 years36 years01.11.1989 to 31.10.2000
Deputy Manager (Data Scientist)24 years32 years01.11.1993 to 31.10.2002

Date-of-birth ranges are our calculation from the age limits for quick self-checking; the notification states only the ages.

Upper age relaxation

CategoryRelaxation
OBC (Non-Creamy Layer)3 years
SC / ST5 years
PwBD (UR / EWS)10 years
PwBD (OBC)13 years
PwBD (SC / ST)15 years

Relaxations are not cumulative. OBC candidates in the creamy layer must apply as UR. OBC candidates need a non-creamy-layer certificate issued between 01.04.2026 and the date of interview; EWS candidates need an Income & Asset Certificate based on FY 2025-26 income.

SBI Data Scientist Educational Qualification

The basic qualification is the same for all three posts (as on 31.10.2026):

Accepted degrees

  • B.E. / B.Tech / M.Tech in Computer Science, IT, Electronics, Electrical & Electronics, Electronics & Communication, Data Science, or AI & ML (or equivalent degree in these disciplines)
  • M.Sc Data Science
  • M.Sc (Statistics / Computer Science)
  • MA (Statistics) or M.Stat
  • MCA

Marks requirement

Minimum 60% marks or equivalent grade in both Graduation and Post Graduation, from a recognised University or Institution.

If your degree certificate does not show the specialisation, you will need a separate certificate from the university that mentions it.

Work Experience Required (Post-Basic Qualification, as on 31.10.2026)

PostMandatory experienceSpecialisation tracks (choose one)
Chief ManagerMinimum 6 years in AI/ML, of which 2+ years in scalable, Python-based LLM / GenAI / Agentic AI applicationsAgentic AI Architecture or AI Quality Assurance and Testing
ManagerOverall 4+ years in AI/ML, of which 2+ years in Python-based LLM / GenAI / Agentic AI solutions at scaleGen AI / Agentic AI, MLOps, LLMOps, AI Security & Governance, DevSecOps, or AI Quality
Deputy Manager3 years overall, of which 2+ years full-time in Analytics / AI / ML / GenAI, and minimum 1+ year full-time developing AI/ML/Deep Learning/GenAI modelsFront end or Integration (EIS / MuleSoft)
Important: training and teaching experience is not counted. Your experience certificate must clearly show the nature of duties, dates and duration, level or position, and responsibilities. If you cannot get such a certificate, other documents that clearly show the same details may be accepted at SBI's discretion. Government and PSU employees need a No Objection Certificate at interview.

Preferred experience by track

Chief Manager: Agentic AI Architecture or AI Quality Assurance

Agentic AI Architecture

  • Defining end-to-end architecture for production-grade LLM, RAG and agentic AI solutions
  • Designing secure solutions for on-premises, private-cloud or controlled enterprise environments
  • Architecture governance, technology evaluation, scalability, security and production readiness
  • Domain experience in banking, document intelligence, risk and compliance, lending, workflow automation or knowledge management

AI Quality Assurance and Testing

  • Leading teams for functional, integration, automation, performance, security and reliability testing
  • Defining test strategies, quality gates, entry and exit criteria, defect governance and release-readiness standards
  • Testing AI/ML, Generative AI and Agentic AI applications, including model and application behaviour
  • Leading teams engaged in Generative or Agentic AI application testing
Manager: six possible tracks
  • Gen AI / Agentic AI: deploying AI apps in enterprise, on-prem or private-cloud environments; exposure to conversational banking, document intelligence, risk and fraud, lending, workflow automation.
  • MLOps: model versioning, release, monitoring, rollback, recalibration, retirement; CI/CD and MLOps platforms.
  • LLMOps: model, prompt, config and evaluation versioning; automated evaluation, guardrails, observability, incident management.
  • AI Security & Governance: AI/ML/LLM/agent security and risk assessment; threat modelling, control testing, incident handling, audit support.
  • DevSecOps: secure CI/CD pipelines; containers, Kubernetes, IaC, vulnerability and release management.
  • AI Quality: testing AI/ML/GenAI/agentic systems; test plans, cases, data, traceability and validation evidence.
Deputy Manager: Front end or Integration
  • Front end: modern web apps with JavaScript/TypeScript frameworks; integrating REST APIs, authentication services and design systems; responsive design, accessibility, performance and secure coding; dashboards and decision-support interfaces.
  • Integration (EIS / MuleSoft): designing, building and supporting enterprise integrations via APIs or integration platforms; connecting core systems, data platforms, digital channels and AI/ML services; secure synchronous, asynchronous and event-driven interfaces.

Preferred Certifications and Specific Skills

PostPreferred certifications
Chief ManagerAI Architecture: TOGAF or equivalent, or AI Quality and Testing: ISTQB or equivalent
ManagerCertifications in Gen AI / Agentic AI, MLOps, LLMOps, AI Security & Governance, DevSecOps, or AI Quality & Testing from recognised universities or institutions
Deputy ManagerFlutter / Angular / React or equivalent, or MuleSoft or equivalent

Skill keywords SBI lists in the notification

These are worth mirroring (truthfully) in your resume and biodata, because shortlisting and the interview will look for them.

Foundation modelsPrompt engineeringStructured outputsFunction callingRAGEmbeddingsVector databasesRerankingKnowledge graphsAgent orchestrationMCPMemory and planningHuman-in-the-loopModel servingGuardrailsObservabilityInference optimisationKubernetesCI/CDMLOpsLLMOpsModel registryFeature storeRed teamingPrompt injection defenceHallucination controlGroundedness evaluationResponsible AIGPU stackReact / AngularREST, SOAP, GraphQLEvent-driven integrationAPI gateways

Chief Manager specific skills (full list)

GenAI / Agentic AI Architecture

  • Foundation models, prompt engineering, structured outputs, function calling
  • RAG: embeddings, vector DBs, reranking, enterprise search, knowledge graphs
  • Agentic AI: orchestration, tools, APIs, MCP, planning, memory, HITL
  • Model serving, caching, guardrails, observability, inference optimisation
  • Microservices, APIs, containerisation, Kubernetes, CI/CD
  • MLOps and LLMOps: versioning, deployment, rollback, monitoring
  • Architecture for conversational AI, document intelligence, assistants, decision support
  • Responsible AI, security, privacy, governance, GPU stack, scalability

AI Quality Assurance and Testing

  • Testing web, mobile, API, microservices and data-pipeline systems
  • GenAI/LLM validation: groundedness, hallucination, toxicity, prompt injection, leakage
  • Agentic AI testing: task completion, planning, permissions, memory, failure recovery
  • Test automation frameworks, CI/CD, synthetic data, regression suites
  • Performance, scalability, resilience, security and DR testing
  • Banking domain validation: reports, dashboards, regulatory controls; UAT and release assurance
Manager specific skills (full list by track)
  • Gen AI / Agentic AI: Python AI apps, APIs, microservices, async processing; foundation models, prompt engineering, structured outputs, function calling, model evaluation; RAG with embeddings, vector DBs, reranking and enterprise search; agents with orchestration, tools, MCP, memory, planning and human oversight; model serving, gateways, caching, guardrails; containers, Kubernetes, CI/CD, MLOps/LLMOps; controlling hallucination, prompt injection, leakage, toxicity and bias.
  • MLOps: CI/CD and DevSecOps for testing, packaging, deployment, promotion and rollback; model registries, feature stores, artefact repositories, experiment tracking.
  • LLMOps: LLM serving, model gateways, inference APIs, caching, routing and fallback; prompt/model/config version control; automated evaluation of groundedness, relevance, safety, bias, task completion; RAG operations from ingestion to knowledge freshness.
  • AI Security & Governance: AI security assessments, threat modelling, architecture reviews; protecting LLM, RAG and agents from injection, leakage, unsafe tools and supply-chain risks; red teaming and adversarial testing; model inventories and risk tiering; AI risk, privacy, compliance and third-party due diligence.
  • DevSecOps: CI/CD, source control, artefact management, release orchestration and rollback; secrets, certificates, identities, code signing and keys; securing containers, Kubernetes and registries; infrastructure and policy as code.
  • AI Quality: testing web, mobile, API, microservices, data-pipeline and event-driven apps; evaluating GenAI for groundedness, relevance, hallucination, toxicity, injection and leakage; validating agent planning, tool selection, permissions, memory and recovery; automation frameworks and CI/CD quality gates.
Deputy Manager specific skills (full list by track)
  • Front end: HTML5, CSS3, JavaScript, TypeScript, responsive UI; React, Angular or similar; state management, routing, validation, reusable components; AI-powered interfaces, conversational assistants and workflow UIs; dashboards, alerts, document upload and case-management flows; rendering, caching and performance; accessibility, localisation, mobile-first; CI/CD and front-end observability.
  • Integration (EIS / MuleSoft): MuleSoft, EIS or equivalent; REST, SOAP, GraphQL, microservices, webhooks, SOA; messaging, queues, streaming, publish-subscribe; data mapping, transformation and routing; API gateways, throttling, versioning and lifecycle; integrating AI services, model APIs and agentic workflows; enterprise integration patterns, API-led connectivity and distributed tracing.

Job Profile, Roles and KRAs

Chief Manager (Data Scientist)

AI Architecture: set end-to-end architecture, reusable patterns and governance for LLM, RAG, conversational AI, document intelligence and agentic workflows; architect model gateways, orchestration layers, vector stores and serving with identity, audit and rollback controls; establish LLMOps; ensure GPU utilisation, resilience and cost efficiency.

KRA: a scalable, vendor-loosely-coupled AI architecture and roadmap sized for several lakh users.

Or AI Quality and Testing: QA strategy, risk-based test planning, release readiness; testing ML, GenAI and agents for accuracy, groundedness, safety, bias and latency. KRA: test coverage, defect leakage, automation ratio, release readiness.

Manager (Data Scientist)

Design, build, deploy and operate secure, scalable AI, ML, GenAI, LLM and agentic-AI solutions for enterprise banking: RAG pipelines, conversational systems, document intelligence, inference APIs, agents and knowledge bases.

KRAs by track: GenAI agents and RAG apps for process automation; MLOps pipelines on SBI's on-premises AI platform; full LLM and agent lifecycle with drift and hallucination monitoring; AI security, red teaming and compliance with RBI frameworks; DevSecMLOps; end-to-end AI quality before production.

Deputy Manager (Data Scientist)

Front end: responsive, accessible interfaces for conversational AI, knowledge assistants, document intelligence, dashboards and agentic workflows, with validation, audit-compliant actions and human-approval flows.

Integration: integrate AI services with core banking systems, APIs and data platforms; build reusable APIs and modular integration components.

SBI notes that actual KRAs are assigned on joining and the listed roles are illustrative.

SBI Data Scientist Salary 2026: Pay Scale by Post

PostGradeBasic pay scale (Rs per month)Starting basic
Chief Manager (Data Scientist)SMGS-IV102300-2980/4-114220-3360/2-120940Rs 1,02,300
Manager (Data Scientist)MMGS-III85920-2680/5-99320-2980/2-105280Rs 85,920
Deputy Manager (Data Scientist)MMGS-II64820-2340/1-67160-2680/10-93960Rs 64,820

On top of basic pay, officers get DA, HRA, CCA, PF, NPS (Contributory Pension Fund), LFC, medical facility, leave and other perks as per SBI's salary structure. Appointment is made at the starting stage of the pay scale.

How to read the scale: for Manager, "85920-2680/5-99320" means basic starts at Rs 85,920 and rises by Rs 2,680 for 5 increments to reach Rs 99,320, then by Rs 2,980 for 2 increments to Rs 1,05,280.

Place of posting: Mumbai, or anywhere in India if administratively required. SBI can also depute officers to associates or subsidiaries.

Starting basic pay of Rs 85,920 for Manager and Rs 64,820 for Deputy Manager, plus DA, HRA, NPS and perks. Give the written test your best shot with the SBI Data Scientist Test Series & Notes.

Get the Test Series & Notes

SBI Data Scientist Selection Process

Chief Manager: Shortlisting + Interview

  • A Bank committee decides the shortlisting parameters. Meeting minimum eligibility does not guarantee a call.
  • Interview of 100 marks; qualifying marks decided by the Bank.
  • Merit list from interview score only; ties broken by age (older ranked higher).

Manager and Deputy Manager: Online Test + Interview

  • Online written test with qualifying General Aptitude and a merit-deciding Professional Knowledge paper.
  • Category-wise shortlisting for interview based on the PK cut-off.
  • Interview of 100 marks.
  • Final merit: Written Test 70% + Interview 30% after normalisation.
  • SBI may drop the written test and use shortlisting + interview instead.

A post-wise and category-wise wait list is kept for one year from the final result (or until a fresh advertisement for the same post, whichever is earlier).

SBI Data Scientist Exam Pattern 2026 (Manager and Deputy Manager)

PaperSectionQuestionsMarksTimeNature
General AptitudeTest of Reasoning151545 min (combined)Qualifying only
Quantitative Aptitude1515
English Language2020
Professional KnowledgeAI/ML domain knowledge6010075 minCounted for merit
Total110150120 minNo negative marking
What this means for you: the 60 PK questions carry 100 marks, so many questions are worth about 1.67 marks each. With no negative marking, attempt every question. General Aptitude only needs to be cleared; spend most of your hours on PK.

Professional Knowledge: Manager (Data Scientist)

#SubjectQuestionsMarksWeight
iProgramming, Data Structures & Machine Learning Fundamentals101717%
iiGenerative AI, LLMs, RAG & Agentic AI122020%
iiiMLOps, LLMOps & DevSecOps122020%
ivAI Security, Governance & Responsible AI91515%
vSoftware Testing & AI Quality Engineering71111%
viCloud, Infrastructure, Databases, Networking & Security Basics101717%
Total60100100%

Professional Knowledge: Deputy Manager (Data Scientist)

#SubjectQuestionsMarksWeight
iProgramming, Data Science & Machine Learning Fundamentals152525%
iiGenerative AI, RAG & Agentic AI Fundamentals203333%
iiiApplication Development, APIs & Integration Basics152525%
ivSecurity, Testing, Cloud & Deployment Basics101717%
Total60100100%

The PK pattern and syllabus are marked tentative in the notification, and SBI says questions may come from outside the listed syllabus.

SBI Data Scientist Syllabus 2026 (Indicative)

(A) Manager (Data Scientist) syllabus

1. Programming, Data Structures & Machine Learning Fundamentals (17 marks)
  • Programming and Data Structures: Python basics; OOP (abstraction, encapsulation, inheritance, polymorphism); arrays, linked lists, stacks, queues, trees, heaps, hashing, recursion; searching and sorting algorithms with time and space complexity.
  • Statistics and ML: basic probability and statistics, hypothesis testing; supervised, unsupervised and reinforcement learning; regression, classification, clustering; overfitting, bias-variance, feature engineering; metrics (accuracy, precision, recall, F1, ROC-AUC); neural networks, CNN, RNN, LSTM; PyTorch and TensorFlow.
2. Generative AI, LLMs, RAG & Agentic AI (20 marks)
  • GenAI and LLMs: Transformer architecture, attention, tokenisation, embeddings; pre-training, fine-tuning (LoRA, QLoRA), RLHF, quantisation; prompt engineering (zero-shot, few-shot, chain-of-thought); hallucination.
  • RAG and Agentic AI: chunking, vector databases, similarity search; agents with planning, memory, tool calling, multi-agent systems, MCP; LangChain, LangGraph, LlamaIndex (basic awareness).
  • Application development: HTTP/HTTPS, REST APIs (GET, POST, PUT, DELETE), JSON; Git; monolith vs microservices.
3. MLOps, LLMOps & DevSecOps (20 marks)
  • MLOps and LLMOps: ML lifecycle, experiment tracking, model registry, feature store (MLflow and similar); pipeline orchestration; batch and real-time deployment; canary and blue-green deployment; data drift, concept drift, performance decay, retraining; prompt versioning, model serving, latency and cost management, guardrails.
  • DevOps and DevSecOps: CI/CD, Git workflows, Jenkins, GitHub Actions; Docker, Kubernetes; Terraform, Ansible; shift-left security: SAST, DAST, dependency scanning, secrets management, container security.
4. AI Security, Governance & Responsible AI (15 marks)
  • AI Security: adversarial attacks, data poisoning, model theft, model inversion; prompt injection, jailbreaking, data leakage; OWASP Top 10 for LLM Applications; guardrails, red teaming, access control, PII masking.
  • Governance and Responsible AI: fairness, bias, transparency, explainability (SHAP, LIME), accountability; model risk management and validation; NIST AI RMF, EU AI Act, ISO/IEC 42001, DPDP Act 2023, RBI FREE-AI guidelines.
5. Software Testing & AI Quality Engineering (11 marks)
  • Software testing: SDLC and STLC; black box, white box, unit, integration, regression testing and UAT; Selenium, PyTest, Postman, JMeter; defect life cycle.
  • Testing AI/ML/LLM apps: data quality testing, model validation, bias and robustness testing; BLEU, ROUGE, LLM-as-a-Judge, golden datasets, human evaluation; RAG evaluation (faithfulness, relevance, groundedness); agent testing.
6. Cloud, Infrastructure, Databases, Networking & Security Basics (17 marks)
  • Cloud: cloud characteristics; SaaS, PaaS, IaaS; public vs private cloud; virtualisation; VMs vs containers; Type 1 vs Type 2 hypervisors.
  • Databases: RDBMS, ER diagrams, ACID transactions, keys, indexes, normalisation, joins; DDL, DML, TCL; SQL functions; views, triggers, cursors; NoSQL and vector databases.
  • OS and Networking: processes, threads, deadlock, memory management, CPU scheduling; Windows and Unix/Linux; OSI and TCP/IP, TCP, UDP, IPv4 vs IPv6, DNS, HTTP, network types.
  • Security: XSS, CSRF, injection, DDoS, brute force; OWASP Top 10; symmetric and asymmetric keys, public-private key encryption, digital signatures; firewalls, access control, remote access VPN; backup and recovery.

(B) Deputy Manager (Data Scientist) syllabus

1. Programming, Data Science & ML Fundamentals (25 marks)
  • Programming: data types, operators, control structures, functions, exception handling; OOP and basic data structures; Python and common data/AI libraries; Git.
  • Data and databases: preprocessing, data quality, feature preparation, EDA; RDBMS, ACID, keys, indexes, normalisation, joins; SQL basics; NoSQL and vector databases; batch and real-time processing.
  • ML and DL: supervised and unsupervised learning; regression, classification, clustering; train/validation/test, cross-validation, overfitting, regularisation; accuracy, precision, recall, F1, ROC-AUC, RMSE; overview of neural networks, transformers, tokenisation and embeddings.
2. Generative AI, RAG & Agentic AI Fundamentals (33 marks)
  • GenAI and LLMs: prompts, context windows, parameters, model responses; zero-shot, few-shot and structured output; hallucination, non-determinism, limitations, evaluation basics; open-source vs proprietary models; small language models.
  • RAG: architecture and flow; ingestion, chunking, embeddings, vector search, retrieval, reranking; groundedness, relevance, faithfulness, knowledge freshness.
  • Agentic AI: planning, memory, tool/function calling; human-in-the-loop, task completion, guardrails; Model Context Protocol (hosts, clients, servers, tools, resources); banking use cases.
  • Responsible AI: fairness, transparency, explainability, privacy, safety, human oversight; bias, toxicity, sensitive-information leakage.
3. Application Development, APIs & Integration Basics (25 marks)
  • App development: client-server and web concepts; monolithic, layered and microservices architectures; HTTP/HTTPS, JSON, XML, serialisation; synchronous, asynchronous and streaming communication; validation, error handling, logging, caching, monitoring.
  • APIs and integration: REST and SOAP, methods, status codes; API design, documentation, versioning, authentication, rate limiting; API gateways, SOA, message queues; transformation, routing, orchestration, event-driven integration; integrating AI/ML services via APIs.
  • User interaction: UI/UX basics; responsive design, accessibility, secure input handling; presenting AI responses, citations, alerts and human-approval steps.
4. Security, Testing, Cloud & Deployment Basics (17 marks)
  • Security: authentication and authorisation; IAM, OAuth 2.0, tokens, SSO; encryption, hashing, certificates, secrets management; injection, XSS, CSRF, insecure access control; prompt injection, data leakage, insecure tool use, excessive agency; least privilege, audit logging.
  • Testing: STLC; unit, functional, integration, API, regression and UAT; test cases, test data, defect lifecycle, basic performance testing; testing GenAI outputs for relevance, groundedness, safety and non-determinism.
  • Cloud and deployment: public, private, hybrid; IaaS, PaaS, SaaS; VMs, containers, Kubernetes; CI/CD; availability, scalability, observability, backup and recovery.

SBI Data Scientist Exam Centres 2026 (Tentative)

The online test may be held at: Guwahati, Silchar, Patna, Mohali, Raipur, Delhi / New Delhi / NCR, Faridabad, Ghaziabad, Greater Noida, Gurugram, Mapusa, Ahmedabad / Gandhinagar, Vadodara / Anand, Ambala, Hamirpur (HP), Jammu, Srinagar, Shimla, Bilaspur (HP), Jamshedpur, Ranchi, Bengaluru, Hubli / Dharwad, Mangalore, Ernakulam, Thiruvananthapuram, Bhopal, Indore, Mumbai / Thane / Navi Mumbai / MMR, Nagpur, Pune, Imphal, Shillong, Aizawl, Kohima, Bhubaneswar, Puducherry, Jaipur, Udaipur, Gangtok, Chennai, Madurai, Hyderabad, Agartala, Lucknow, Kolkata and Siliguri.

Your centre choice cannot be changed later, and SBI may allot a different centre.

Application Fee

CategoryFee and intimation charges
General / EWS / OBCRs 750
SC / ST / PwBDNil

Pay online by debit card, credit card, internet banking or UPI. Transaction charges are borne by the candidate. The fee is non-refundable and cannot be adjusted against any other recruitment.

How to Apply for SBI Data Scientist Recruitment 2026

  1. Go to sbi.bank.in/web/careers/current-openings, scroll to advertisement CRPD/SCO/2026-27/31 and read the full notification.
  2. Prepare scans: photo, signature, left-hand thumb impression, hand-written declaration, plus all PDFs listed below (biodata, resume, ID, DOB proof, marksheets, experience certificates, etc.).
  3. Click Apply Online and register with a valid e-mail ID and mobile number. Note the provisional registration number and password.
  4. Fill the form carefully. You can save and edit, but editing is allowed only three times.
  5. Upload photo, signature, thumb impression, declaration and documents, and capture a live photograph via webcam or by scanning a QR code with your phone.
  6. Check every field and preview each upload. No change is allowed after final submission.
  7. Pay the fee (if applicable). Registration is complete only after successful payment.
  8. Download and keep the e-receipt and the system-generated application form.
Apply for only one post. SBI does not allow applications to more than one post. If you apply more than once, only the last completed application is kept and other fees are forfeited.

Applied? Now prepare. The SBI Data Scientist Test Series & Notes covers both the Manager and Deputy Manager papers.

Enroll now

Documents and Upload Specifications

Documents to upload (PDF, A4, max 500 KB each)

  1. Recent photograph and signature (JPG)
  2. Biodata (format attached to the notification)
  3. Resume
  4. ID proof
  5. Proof of date of birth
  6. Educational certificates: relevant mark-sheets and degree certificate
  7. Experience certificates
  8. Caste / EWS certificate (if applicable)
  9. PwBD certificate (if applicable)
  10. Preferred qualification / certification (if any)
  11. Form-16 / offer letter / latest salary slip from current employer

Image specifications

ItemFormatDimensions (preferred)File size
Photograph (passport style, colour, light background)JPG / JPEG200 x 230 px20 to 50 KB
Signature (black ink on white paper, not in capital letters)JPG / JPEG140 x 60 px10 to 20 KB
Left-hand thumb impression (black or blue ink on white paper)JPG / JPEG240 x 240 px at 200 DPI (3 cm x 3 cm)20 to 50 KB
Hand-written declaration (English, black ink, not in capital letters)JPG / JPEG800 x 400 px at 200 DPI (10 cm x 5 cm)50 to 100 KB
All other documentsPDF (A4)Scan at minimum 200 DPIUp to 500 KB

Text of the hand-written declaration

"I, ________ (Name of the candidate), ________ (Date of Birth) hereby declare that all the information submitted by me in the application form is correct, true and valid. I will present the supporting documents as and when required. The signature, photograph and left-hand thumb impression is of mine."

If you do not have a left thumb, you may use the right thumb.

Key Rules to Remember

  • Eligibility date: all criteria are checked as on 31.10.2026; admission at every stage is provisional until document verification.
  • Tie-break: candidates with equal marks at the cut-off are ranked by age, older first.
  • EWS: Income & Asset Certificate based on FY 2025-26 income is required at document verification.
  • Exam day: carry the call letter and an original photo ID (Passport, Aadhaar, PAN, Driving Licence, Voter ID, or bank passbook with attested photo) plus a self-attested photocopy. Mobile phones, smartwatches and calculators are not allowed. Left-eye IRIS scan and biometric checks may be done.
  • Travel reimbursement for interview: economy air fare by the shortest route, up to Rs 15,000 (both ways) for Chief Manager and up to Rs 10,000 for Manager and Deputy Manager, or actual fare if lower.
  • Service: selected officers serve probation as per SBI policy and can be posted or transferred anywhere in India or deputed to associates or subsidiaries.
  • Jurisdiction: any dispute falls only under courts in Mumbai.
  • Data consent: applying means consenting to processing of your personal data under the DPDP Act, 2023.

How to Prepare for the SBI Data Scientist Exam 2026

With the test expected in December 2026 or January 2027, you have roughly 4 to 8 weeks. Here is the plan I recommend to working professionals preparing for this exam. It is based on the marks weight in the official pattern above.

The fastest way to follow this plan is with the SBI Data Scientist Test Series & Notes: study the notes week by week and use the mocks to check your progress.

Week-wise plan (6 weeks)

WeekManager focusDeputy Manager focus
1Python, OOP, DSA complexity; probability, hypothesis testing; ML metricsPython, OOP, SQL; EDA, preprocessing; ML metrics including RMSE
2Transformers, attention, embeddings; LoRA/QLoRA, RLHF, quantisation; promptingLLM basics, context windows, structured output, hallucination; RAG flow end to end
3RAG, vector DBs, agents, MCP, LangGraph; REST and GitAgents, tool calling, MCP hosts/clients/servers; HITL and guardrails
4MLflow, model registry, drift, canary vs blue-green; CI/CD, Docker, Kubernetes, Terraform; SAST vs DASTREST vs SOAP, status codes, API gateways, rate limiting, message queues, event-driven patterns
5OWASP LLM Top 10, adversarial attacks, SHAP/LIME, NIST AI RMF, EU AI Act, ISO 42001, DPDP, RBI FREE-AI; testing tools and RAG evaluationOAuth 2.0, SSO, hashing vs encryption, XSS/CSRF; STLC and GenAI output testing; cloud models and Kubernetes basics
6Cloud, DBMS, OS, networking revision; daily full-length mocks; one Reasoning/Quant/English sectional per day to stay above qualifying marks

High-yield topics most aspirants skip

  • Regulatory frameworks: NIST AI RMF functions (Govern, Map, Measure, Manage), EU AI Act risk tiers, ISO/IEC 42001 as an AI management system standard, DPDP Act 2023 basics and RBI's FREE-AI framework. These are easy marks in the Manager paper.
  • LLM evaluation vocabulary: faithfulness vs relevance vs groundedness, BLEU vs ROUGE, LLM-as-a-Judge, golden datasets.
  • Deployment strategies: canary vs blue-green, batch vs real-time inference, data drift vs concept drift.
  • MCP basics: what hosts, clients, servers, tools and resources are. This is named explicitly in both syllabi.
  • Classic IT basics: ACID, normalisation, deadlock, CPU scheduling, OSI layers, Type 1 vs Type 2 hypervisors. These repeat across every bank IT officer exam.

All these high-yield topics are covered with exam-style questions in the SBI Data Scientist Test Series & Notes.

Get the Test Series & Notes

Interview and shortlisting tips

  • Choose one track (for example LLMOps or AI Security) and make sure your resume, biodata and experience certificate tell the same story.
  • Be ready to explain one production GenAI system you built: architecture, evaluation, guardrails, cost and failure handling.
  • Know banking AI use cases SBI names: conversational banking, document intelligence, risk and fraud, lending and workflow automation.

SBI Data Scientist Recruitment 2026: FAQs

Which is the best test series for SBI Data Scientist 2026?

The SBI Data Scientist Test Series & Notes by PiyushAI is built specifically for Advt. CRPD/SCO/2026-27/31. Its notes and mocks follow the official Professional Knowledge pattern for both Manager and Deputy Manager (Data Scientist).

How should I prepare for the SBI Manager (Data Scientist) exam?

Spend most of your time on the 100-mark Professional Knowledge paper, especially GenAI/RAG/Agentic AI and MLOps/LLMOps/DevSecOps (20 marks each). Revise from exam-focused notes and take timed mocks; the SBI Data Scientist Test Series & Notes provides both.

How many vacancies are there in SBI Data Scientist Recruitment 2026?

20 in total: Chief Manager 2, Manager 14 and Deputy Manager 4. Category-wise: SC 3, ST 1, OBC 2, EWS 1, UR 13.

What is the last date to apply?

10 December 2026. The window opened on 20 November 2026.

What is the age limit?

As on 31.10.2026: Chief Manager 27 to 37, Manager 26 to 36, Deputy Manager 24 to 32 years, with relaxation for OBC (3), SC/ST (5) and PwBD (10 to 15).

What is the exam pattern?

General Aptitude (Reasoning 15, Quant 15, English 20; 45 minutes; qualifying) and Professional Knowledge (60 questions, 100 marks, 75 minutes). No negative marking. Final merit: written 70%, interview 30%.

Is there a written exam for Chief Manager?

No. Chief Manager is selected through shortlisting and a 100-mark interview.

Is there negative marking?

No, the notification clearly says there are no negative marks for wrong answers.

What is the salary of SBI Deputy Manager (Data Scientist)?

Basic pay starts at Rs 64,820 in MMGS-II (scale up to Rs 93,960), plus DA, HRA, CCA, PF, NPS, LFC, medical and other benefits.

Can freshers or final-year students apply?

No. Every post needs post-qualification work experience as on 31.10.2026 (3, 4 or 6 years).

Does teaching or training experience count?

No. SBI states training and teaching experience will not be counted for eligibility.

Can MCA or M.Sc Statistics graduates apply?

Yes. MCA, M.Sc Data Science, M.Sc Statistics/Computer Science, MA Statistics and M.Stat are accepted, with 60% in graduation and post-graduation.

Can I apply for both Manager and Deputy Manager?

No. Only one post is allowed. If you apply to more than one, only the last completed application is retained.

Where will selected candidates be posted?

Mumbai, or anywhere in India as per administrative need.

Is this page official?

No. This is an independent explainer by PiyushAI. Always confirm details in the official notification on the SBI careers website.

SBI Data Scientist 2026: notes + mock tests for Manager and Deputy Manager, by Piyush Wairale (IIT Madras).

Enroll now