New opportunity

Lead Data & AI Engineer

Qualys · Pune, India

About this role

Lead the design and delivery of enterprise-grade data and AI solutions for Qualys’s CIO organization in Pune, India. The role focuses on building an AI-ready data lake or lakehouse from scratch and using it to support process automation, decision-making and business applications.

Responsibilities

Architect ingestion, storage, transformation, serving and governance layers on a cloud data platform. Implement Bronze/Silver/Gold architecture for structured, semi-structured and unstructured data, and build reliable batch and streaming pipelines from enterprise applications and product systems.
Establish data quality frameworks, data contracts, SLAs, lineage and metadata management. Design role-based, row-level and column-level access controls across enterprise sources. Evaluate data lake platforms and recommend an option based on cost, governance, AI readiness and operational overhead.
Build AI applications including natural-language access to enterprise data, RAG-based search, automated reporting and workflow agents. Integrate LLM APIs such as OpenAI, Anthropic Claude or AWS Bedrock into production workflows. Assess AI use cases and build-versus-buy choices, advise leadership, collaborate with business teams, develop reusable standards and mentor junior engineers.

Required qualifications

7–12+ years of hands-on data engineering experience, including at least 2–3 years building enterprise AI solutions. Experience with enterprise data architectures, lakehouses, semantic layers, governance and AI-ready data foundations.
Deep expertise in Snowflake, Databricks or an equivalent modern platform; strong Python, backend development, API, integration, orchestration and SQL skills. Production GenAI or LLM application experience involving RAG, LLM APIs, agents, NL-to-SQL or evaluation frameworks.
Hands-on AWS, Azure or GCP infrastructure experience; batch or streaming pipelines at multi-TB daily scale or higher; knowledge of dimensional modelling, medallion architecture or data vault, and enterprise access control and lineage. Ability to explain technical strategy and investment cases to leadership.
BE/B.Tech/MCA, preferably in Computer Science, Information Technology or a related field.

Preferred qualifications

Enterprise application integration, reusable AI frameworks, vector databases, embeddings, semantic search, advanced RAG, real-time streaming, dbt, data cataloguing and lineage tools. Product, SaaS, cybersecurity or high-growth technology experience is also preferred. The position is full-time.

For full details about this role, click “Apply on company site” to view the employer’s original posting.

Skills for this role

Data engineeringData lake architectureLakehouse architectureSnowflakeDatabricksDelta LakeApache IcebergPythonSQLData pipelinesBatch processingStream processingMedallion architectureData modellingDimensional modellingData vaultData governanceData qualityData contractsData lineageMetadata managementRBACRow-level securityColumn-level securityAPIsSystem integrationWorkflow orchestrationGenerative AILarge language modelsRetrieval-augmented generation (RAG)

Checking your profile…

Preparation suggestions, not the employer’s actual interview questions. Always check the original posting for current requirements.

Understand the role

What matters most

Build the AI-ready lakehouse foundation
The role owns a cloud data lake or lakehouse from scratch, spanning ingestion, storage, transformation, serving and governance. Its Bronze/Silver/Gold design must handle structured, semi-structured and unstructured data across business domains.
Deliver dependable data pipelines
Build scalable batch and streaming pipelines from enterprise applications and product systems. The posting also calls for data quality frameworks, contracts and SLAs across ingestion and transformation, with experience at multi-TB daily scale.
Make enterprise data trusted and controlled
Establish lineage and metadata so assets are discoverable and auditable. Design role-based, row-level and column-level controls while harmonizing identities and entitlements across source systems.
Put AI into production workflows
Build applications such as natural-language data access, RAG search, automated reporting and workflow agents on the data platform. Integrate LLM APIs into production workflows and business-facing applications.
Set technical direction and enable others
Own the data-platform evaluation across cost, governance, AI readiness and operational overhead. Advise leadership on AI use cases and build-versus-buy choices, collaborate with business teams, establish reusable standards and mentor junior engineers.

Likely challenges to think through

Choosing a platform without optimizing for one dimension
Likely preparation challenge, not a stated employer problem: the lowest-cost option may differ from the easiest to govern or operate for AI workloads. Prepare a decision framework comparing plausible platforms against the four evaluation criteria named in the posting, with explicit assumptions and trade-offs.
Balancing ingestion speed with trustworthy data
Likely preparation challenge: integrating varied enterprise and product sources at scale can put delivery speed and freshness in tension with contracts, quality checks and SLAs. Prepare a phased pipeline design showing where to validate data, handle failures and measure reliability for batch and streaming paths.
Preserving source permissions in shared data and AI access
Likely preparation challenge: harmonizing entitlements across systems becomes harder when data is served through natural-language interfaces or RAG. Prepare to explain how role-, row- and column-level controls would carry through retrieval and serving, and how lineage would support auditing.
Selecting AI use cases that justify production investment
Likely preparation challenge: a compelling prototype may not justify its operating cost or meet the reliability needs of a business workflow. Prepare a use-case assessment that weighs measurable outcomes, required data foundations, evaluation of outputs and build-versus-buy options for leadership.

Practice your answers

Technical practice

How would you design a cloud lakehouse from scratch for multiple enterprise data domains?

What to coverSuggested preparation: Explain the ingestion, storage, transformation, serving and governance layers, and what belongs in Bronze, Silver and Gold for structured, semi-structured and unstructured data. Discuss how you would make the foundation usable for both analytics and AI applications.

How would you make batch and streaming ingestion reliable at multi-TB daily scale?

What to coverSuggested preparation: Compare batch and streaming for enterprise applications and product sources, including freshness and operational cost. Explain how you would handle retries, duplicate records, schema changes and recovery while meeting pipeline SLAs.

How would you enforce data contracts and quality across ingestion and transformation?

What to coverSuggested preparation: Describe checks at source boundaries and between lakehouse layers, plus how failures would be detected and contained. Address contract changes, ownership and the relationship between quality measures and SLAs.

How would you make data assets discoverable, traceable and auditable?

What to coverSuggested preparation: Explain how metadata and lineage would connect enterprise sources, transformations and served assets. Discuss how you would verify that lineage remains useful when pipelines or schemas change.

How would you reconcile source-system entitlements with lakehouse access controls?

What to coverSuggested preparation: Distinguish role-based access from row-level restrictions and column-level security or masking. Explain how you would test permissions across sources and prevent a new analytical or AI interface from bypassing them.

How would you recommend a data lake or lakehouse platform to leadership?

What to coverSuggested preparation: Compare Snowflake, Databricks or an equivalent platform against the posting's criteria: cost, governance, AI readiness and operational overhead. State what evidence or a limited evaluation would resolve important trade-offs rather than assuming one platform is best.

When would you use dimensional modelling, medallion layers or data vault, and how would you tune the resulting SQL workloads?

What to coverSuggested preparation: Explain the different purposes of these patterns and how a serving model should fit its analytical use case. Discuss query plans, data layout and workload measurements as ways to diagnose poor performance.

How would you build and evaluate RAG-based search over enterprise data?

What to coverSuggested preparation: Cover content preparation, retrieval, embeddings or semantic search where useful, and how retrieved evidence reaches the model. Discuss measures of retrieval and answer quality alongside freshness and access control.

How would you offer natural-language access to enterprise data without allowing unsafe queries?

What to coverSuggested preparation: Explain the role of a semantic layer in mapping business terms to governed data and how you would validate generated SQL. Address user permissions, ambiguous questions, query cost and evaluation of answer correctness.

How would you integrate LLM APIs and workflow agents into production applications?

What to coverSuggested preparation: Describe the Python, backend API and orchestration boundaries between data workflows, a model provider and a business application. Compare automation with human review, and explain how you would monitor failures, quality and operational cost.

Behavioural practice

Tell me about a time you made a technical investment case to non-technical leaders.

What to coverSuggested practice: Use a real situation, your action and the result. Explain how you presented alternatives, costs and expected business value, including any uncertainty in the recommendation.

Describe a time you worked with business and application teams to define a data or AI use case.

What to coverSuggested practice: Give a real situation, your actions to reconcile different needs and the result. Focus on how you established a useful outcome and identified the data needed to support it.

Tell me about a time you helped other engineers adopt a reusable standard or framework.

What to coverSuggested practice: Describe the situation, how you mentored colleagues or improved the standard, and the result. Include how you handled feedback or differing engineering approaches.

Describe a time a data quality or pipeline issue threatened a downstream commitment.

What to coverSuggested practice: Walk through the real situation, your diagnosis and communication, and the result. Explain how you balanced immediate recovery with a lasting improvement to reliability.

Tell me about a time you chose between building a capability and using an existing option.

What to coverSuggested practice: Use a real situation, the criteria you applied and the outcome of your decision. Discuss trade-offs such as delivery time, maintainability and fit for the intended users without claiming there was a universally correct choice.

Sharpen your knowledge

Topics to prepare

Lakehouse architecture and modelling
The role calls for a lakehouse built from scratch, with ingestion, storage, transformation, serving and governance layers, plus Bronze/Silver/Gold data across multiple domains. To revise, sketch those layers for structured, semi-structured and unstructured sources, then explain when you would choose dimensional modelling or data vault for a serving use case.
Data trust and enterprise access
The posting explicitly requires data quality, contracts, SLAs, lineage, metadata and RBAC with row- and column-level controls. To revise, trace one data asset from source to serving layer and specify its quality checks, contract, ownership metadata, lineage and access rules.
Pipeline reliability at scale
Batch and streaming ingestion from enterprise applications and product systems, including multi-TB daily scale, is stated. To revise, work through how a pipeline would handle late data, retries, schema changes and an SLA breach without silently corrupting downstream data.
Production AI on enterprise data
The role includes natural-language data access, RAG search, reporting, workflow agents and production LLM integration. To revise, diagram a RAG or NL-to-SQL request path and explain how data permissions, answer evaluation and failure handling would apply.
Platform and investment trade-offs
The lead is expected to recommend a data lake platform on cost, governance, AI readiness and operational overhead, and advise on AI build-versus-buy choices. To revise, prepare a short decision matrix comparing Snowflake, Databricks or an equivalent against those stated criteria, with assumptions made explicit.

Skills to revise

Build a governed ingestion slice
As a practice exercise, use a platform you know from the posting’s options to ingest a sample application feed, transform it through Bronze, Silver and Gold, and add a data-quality check. This rehearses the required hands-on lakehouse and pipeline work; the sample is preparation, not a stated Qualys project.
Tune analytical SQL
Take a large analytical query, inspect its execution plan, revise joins or filtering, and compare runtime and data scanned. The posting specifically calls for SQL query optimization, performance tuning and large-scale analytical workloads.
Implement source-to-serving controls
Practice applying RBAC, a row restriction and column masking to a small multi-source dataset; verify results under different roles and record lineage. These controls and lineage are explicit responsibilities, while the dataset is only a rehearsal.
Prototype and evaluate an LLM workflow
Build a small Python API-backed RAG or NL-to-SQL flow using an LLM API named in the posting, then test permitted versus restricted questions, unsupported answers and API failures. This practices the required backend, integration and production GenAI skills without implying a particular implementation is mandated.
Present a technical recommendation
Draft a one-page platform recommendation with assumptions, cost and operating trade-offs, governance implications and AI readiness; rehearse a brief explanation for non-technical leadership. The posting explicitly assigns platform evaluation and calls for communicating investment cases.

Your preparation checklist

Sketch a lakehouse design
Prepare a hypothetical CIO data-platform diagram covering ingestion, storage, Bronze/Silver/Gold transformation, serving and governance for structured, semi-structured and unstructured data. Explain the modelling choices; this is a practice design, not a stated Qualys architecture.
Walk through a relevant data project
If you have one, prepare a concise walkthrough of your role, cloud platform, Python and SQL work, enterprise integrations, orchestration, batch or streaming scale, reliability decisions and measured outcomes. The posting calls for hands-on delivery at multi-TB daily scale or higher.
Build a platform recommendation
Create a short comparison of Snowflake, Databricks or another platform you know, with explicit assumptions. The role owns a recommendation based on cost, governance, AI readiness and operational overhead; be ready to explain the investment case to leadership.
Demonstrate trust and access controls
Prepare one concrete example or design showing data-quality checks, contracts, SLAs, lineage and metadata, plus RBAC and row- and column-level controls across sources. Mark proposed controls separately from controls you have implemented.
Explain a production AI application
If available, bring a walkthrough of a RAG, natural-language data access or workflow-agent application: data permissions, LLM API integration, evaluation, monitoring and business outcome. Otherwise, prepare a clearly labelled design exercise; these are application types named in the posting.
Prepare a leadership and mentoring example
Select an example, if available, of explaining a technical trade-off to non-technical stakeholders, collaborating with application or business teams, establishing reusable standards or mentoring engineers. State your contribution and the decision or outcome.
Write questions for the employer
Ask which CIO data sources and AI use cases are priorities, what platform constraints already exist, how success will be measured, and who owns identity and governance decisions. Treat these as questions, not assumptions about Qualys’s current environment.

Availability last checked: 5 Nov 2026. Status can change after a check; confirm on the employer’s website. Report an issue · Our checking policy

KEEP EXPLORING

Similar AI jobs

Related skills and specializations in India. Matched to this role, not your profile.

Explore more jobs

Finding similar opportunities…