New opportunity

Knowledge Engineer

Accenture · Chennai, India

About this role

Accenture seeks a Knowledge Engineer to lead knowledge engineering within a large AI program. The role turns expert and unstructured knowledge into ontologies, knowledge graphs and semantic models that help AI systems reason, support automation and generate actionable insights. The position is based at Accenture’s Chennai office and is full-time.

Responsibilities

Set technical direction for scalable AI and knowledge graph solutions; lead the build of graph solutions that transform client data architecture, and direct the design and integration of semantic and AI components.
Design, evaluate and maintain ontologies and knowledge models, establish standards, and apply knowledge graph analytics to business problems.
Lead and mentor a team, make decisions across workstreams, and coordinate end-to-end delivery with engineering, research, product, project and client teams.
Build the business case for semantic-layer solutions, develop client relationships, contribute to sales and pre-sales, and provide thought leadership on technology trends, opportunities and risks.

Qualifications and skills:

The posting requires a minimum of 12 years of experience, proficiency in Microsoft Azure Data Services, and 15 years of full-time education. It also specifies a minimum of 12 years of experience in Microsoft Azure Data Services and lists a bachelor’s degree or equivalent.
It calls for at least four listed qualifications, including experience with knowledge graph technologies such as RDF, SPARQL, LPG and SHACL; ontology and schema design; graph-based ML; AI data pipelines, especially for LLMs; databases; team or workstream leadership; cloud platforms; and Python with ML and ETL tools. The listed thresholds range from two to four years depending on the qualification.
Other listed qualifications include practical NLP or search experience, prompt engineering and enterprise LLM applications, team leadership, cross-functional collaboration across time zones, client-facing consulting, a Ph.D. in a related field, and experience with diverse ML techniques and agentic systems.
Graph databases and data engineering are identified as good-to-have skills. The role also calls for data integration, large-scale data handling and data-quality skills.

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

Skills for this role

Knowledge engineeringKnowledge graphsOntologiesSemantic modelingMicrosoft Azure Data ServicesGraph databasesData engineeringRDFSPARQLLPGSHACLSchema designKnowledge graph curationGraph-based machine learningData pipelinesRelational databasesObject storesVector databasesStardogNeo4jAmazon NeptuneAWSAzureGCPPythonTensorFlowPyTorchApache NiFiAirflowNatural language processing (NLP)

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

Own the knowledge models
The role leads ontology and knowledge-model design, evaluation, maintenance, and standards. It turns expert and unstructured knowledge into contextually accurate graphs and semantic models that support AI reasoning, automation, and insights.
Set scalable technical direction
The posting calls for leading graph solutions that transform client data architecture and integrating AI and semantic components. Microsoft Azure Data Services is a must-have skill; data integration, large-scale handling, and data quality are also specified.
Lead delivery across teams
This is a team-lead role with responsibility for decisions across workstreams, mentoring junior colleagues, and coordinating engineering, research, product, delivery, and client stakeholders.
Make the client case for the semantic layer
The role must justify its recommended solution to business and technology stakeholders, build client relationships, contribute to sales and pre-sales, and communicate technology opportunities and risks.

Likely challenges to think through

Balancing model rigor with scale
A likely challenge is preserving meaning, context, and data quality as knowledge models cover large, varied data sets. Prepare a concrete example of how you would set modeling standards, validate a graph, and decide when added detail is worth its maintenance cost.
Making graph and AI components work together
The posting spans semantic models, graph analytics, cloud data integration, and AI pipelines. A plausible difficulty is choosing where structured knowledge adds value without overcomplicating the architecture. Prepare to explain integration boundaries, evaluation criteria, and trade-offs for an end-to-end design.
Aligning technical choices with client value
Leading delivery while making a business case may require balancing technical ambition against stakeholder priorities and foreseeable risks. Prepare an example of how you would explain a semantic-layer recommendation, its limits, and a practical delivery sequence to both technical and business audiences.

Practice your answers

Technical practice

How would you architect a scalable knowledge graph solution using Microsoft Azure Data Services?

What to coverSuggested preparation: Explain how data ingestion, transformation, semantic modeling and downstream AI components would fit together, without assuming a particular Azure product is required. Discuss scale, integration boundaries and how you would evaluate whether the architecture meets the client's needs.

How would you turn expert knowledge and unstructured information into a knowledge model?

What to coverSuggested preparation: Describe how you would identify concepts, relationships and context, then check the representation with domain experts. Address ambiguity, provenance and the risk of losing meaning during extraction.

How would you design and maintain an ontology used across multiple workstreams?

What to coverSuggested preparation: Cover modeling conventions, reuse, versioning and review standards that other teams can follow. Explain how you would evaluate changes against accuracy, usability and downstream effects.

When would you use RDF or a labeled property graph, and how would SPARQL and SHACL fit?

What to coverSuggested preparation: Compare the modeling and querying implications of RDF and labeled property graphs for a stated use case. Explain where SPARQL queries and SHACL validation would help, while acknowledging that the best choice depends on the data and integration needs.

How would you choose between relational, graph, object and vector stores in a semantic solution?

What to coverSuggested preparation: Explain the distinct roles these stores could play and the trade-offs in access patterns, integration and maintenance. The posting lists all four types; frame any proposed combination as a design choice, not an existing Accenture architecture.

How would you build an end-to-end data pipeline for a knowledge-graph-enabled AI application?

What to coverSuggested preparation: Walk through ingestion, transformation, model updates and delivery to an AI application, including where Python or ETL tooling could help. Discuss how you would make pipeline failures visible and manage changes to upstream data.

How would you protect data quality and contextual accuracy as a knowledge graph scales?

What to coverSuggested preparation: Discuss checks for inconsistent entities, relationships and source context, plus a process for correcting and revalidating data. Explain how you would prioritize quality controls when handling large datasets.

How would you evaluate whether graph analytics or graph-based ML solves a business problem?

What to coverSuggested preparation: Start with the decision or insight needed, then describe an appropriate graph-based approach and a comparison against a simpler alternative. Address data suitability, evaluation measures and whether the result is actionable.

How could a knowledge graph support NLP, search or an enterprise LLM application?

What to coverSuggested preparation: Explain a plausible way to connect extracted concepts or graph context to retrieval, prompts or application responses. Discuss relevance, contextual accuracy and how you would test whether the graph adds value.

How would you assess a proposed semantic layer before recommending it to a client?

What to coverSuggested preparation: Describe the technical evidence you would seek for integration, maintainability and AI usefulness, alongside the business value it should demonstrate. Discuss trade-offs, foreseeable limitations and risks rather than assuming a semantic layer is always warranted.

Behavioural practice

Tell me about a technical decision you led across multiple workstreams.

What to coverSuggested preparation: Use a real situation to explain the competing needs, your decision process and how you coordinated implementation. State the result and what you would change in hindsight.

Describe a time you aligned engineering, research and product teams on a shared objective.

What to coverSuggested preparation: Give the situation, the disagreement or coordination challenge, and the actions you took to establish alignment. Explain the result, including how you handled communication across teams or time zones if relevant to your example.

Tell me about a time you mentored someone on a complex technical problem.

What to coverSuggested preparation: Describe the knowledge gap, the guidance you provided and how you balanced support with ownership. Share an observable result without overstating your contribution.

Describe a time you made a technical recommendation to a client or business stakeholder.

What to coverSuggested preparation: Explain the stakeholder's goal, how you presented value and trade-offs, and what you did to earn trust. State the decision or other result that followed.

Tell me about a time you challenged an emerging technology approach because of its risks.

What to coverSuggested preparation: Set out the opportunity, the limitation you identified and how you communicated an alternative or mitigation. Explain the result and what it taught you about setting technical direction.

Sharpen your knowledge

Topics to prepare

Azure data architecture for knowledge graphs
Suggested revision exercise: sketch how data moves through Microsoft Azure Data Services into a knowledge graph, noting integration, scale and quality decisions without assuming a particular service. Relevance: Azure Data Services is the posting’s must-have skill, and the role sets technical direction for client data architecture.
Ontology and semantic-model design
Suggested revision exercise: take a small business domain and explain its entities, relationships, constraints and approach to maintaining the model as knowledge changes. Relevance: the role designs and maintains ontologies, translates expert knowledge and establishes modeling standards.
Knowledge-graph representations and validation
Suggested revision exercise: compare how RDF and LPG would represent the same relationship, then explain where SPARQL queries and SHACL validation fit. Relevance: the posting explicitly lists RDF, LPG, SPARQL and SHACL among knowledge-graph qualifications.
Graph analytics and graph-based ML
Suggested revision exercise: choose a business question, identify a graph-derived signal that could help answer it, and describe how you would check whether it produces useful insights. Relevance: the role applies graph analytics to business problems and lists graph-based ML experience.
AI pipelines, context and data quality
Suggested revision exercise: trace how unstructured material could become curated knowledge for an LLM application; identify checks for accuracy, context and large-scale data quality. Relevance: the posting calls for AI data pipelines, especially for LLMs, and accurate knowledge representations.
Semantic-layer value and delivery trade-offs
Suggested revision exercise: outline a client-facing case for a semantic layer, including expected value, integration trade-offs and foreseeable risks. Relevance: this manager-level role makes the business case, directs work across teams and contributes to pre-sales.

Skills to revise

Design an Azure-based integration plan
Suggested practice: draw a source-to-graph architecture using Microsoft Azure Data Services and annotate transformation, scale, data-quality and ownership decisions. Relevance: Azure proficiency is required, and the role leads graph solutions that transform client data architecture.
Draft and review an ontology
Suggested practice: create a compact domain model from a short unstructured brief, then review it for ambiguous terms, missing relationships and maintainability. Relevance: the role turns expert and unstructured knowledge into ontologies and leads knowledge-model refinement.
Query and validate graph data
Suggested practice: write a small RDF example, a SPARQL query answering a business question, and a SHACL constraint that catches one data error. Relevance: all three technologies are explicitly listed among knowledge-graph qualifications.
Prototype an AI data pipeline
Suggested practice: use Python to outline or build a repeatable flow from source records through transformation and quality checks to graph-ready data; explain how it could support an LLM application. Relevance: the posting lists Python, ETL tooling and hands-on AI pipeline implementation.
Lead a cross-team design review
Suggested practice: rehearse presenting one graph architecture decision to engineering, product and client stakeholders, documenting alternatives, risks and the decision. Relevance: the role makes team decisions, mentors others and coordinates delivery across technical and client teams.

Your preparation checklist

Map the stated qualifications
Prepare a private evidence sheet for the posting’s minimum 12 years of experience, its separate 12-year Microsoft Azure Data Services statement, 15 years of full-time education, and bachelor’s degree or equivalent. Note which of the listed technical and leadership qualifications you could substantiate; do not assume the criteria are interchangeable.
Build a knowledge-graph project walkthrough
Prepare a diagram and brief account of how expert or unstructured knowledge became an ontology, graph, or semantic model. Show design choices, curation, data-quality checks, and how the result supported reasoning or insights. This is a suggested way to demonstrate the role’s stated work.
Practice an Azure-centered design exercise
Sketch a scalable architecture that integrates source data, transformation, a semantic layer, and AI components using Microsoft Azure Data Services. Be ready to explain integration, large-scale data handling, contextual accuracy, and the trade-offs you would validate rather than claim a prescribed Accenture architecture.
Prepare technical decisions and evidence
Select examples you can discuss for relevant listed areas, such as RDF or LPG, SPARQL, SHACL, ontology management, graph analytics or ML, and AI data pipelines for LLMs. For each, record your specific contribution, a design decision, and how you evaluated the result.
Rehearse a leadership and client walkthrough
Prepare one concise account of setting technical direction, coordinating workstreams, mentoring, and resolving a cross-team decision. Add how you would explain the value, limitations, and risks of a semantic-layer proposal to a client; these activities appear in the posting.
Draft targeted employer questions
Bring questions about the knowledge domains and data sources, the current Azure and graph stack, expected scope of team and client ownership, and success measures for the semantic layer. Ask how the heading’s “10–12 years” aligns with the description’s minimum 12 years and its separate Azure experience statement.

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…