Website:
swastikawebtechnology.com
Job details:
We are looking for a Senior AI Knowledge & Governance Architect to design, organize, and maintain the knowledge and governance systems that support our AI initiatives.
The ideal candidate should understand how LLMs, RAG, knowledge bases, vector databases, prompt systems, AI workflows, and data governance work together. You will be responsible for making AI-generated information more accurate, reliable, traceable, secure, and useful for business teams.
This is a hands-on role for someone who can work across AI knowledge architecture, documentation, governance frameworks, and practical AI implementation.
Key ResponsibilitiesAI Knowledge Architecture
- Design and maintain structured knowledge systems for AI applications.
- Build and organize knowledge bases used by LLM and RAG-based applications.
- Define knowledge structures, taxonomies, metadata, tagging, and information hierarchies.
- Improve the quality, consistency, and accessibility of information used by AI systems.
- Identify knowledge gaps, duplication, outdated information, and conflicting sources.
RAG & LLM Knowledge Systems
- Work with Retrieval-Augmented Generation (RAG) systems and enterprise knowledge repositories.
- Understand document ingestion, chunking, embeddings, vector search, retrieval, and reranking.
- Help improve retrieval accuracy and reduce hallucinations.
- Define strategies for source attribution and knowledge traceability.
- Evaluate AI responses against approved knowledge sources.
AI Governance
- Develop practical governance standards for AI systems and knowledge assets.
- Establish guidelines for data quality, source validation, access control, and information lifecycle management.
- Define processes for reviewing and approving AI knowledge sources.
- Maintain documentation around AI systems, datasets, prompts, models, and knowledge sources.
- Help establish controls around sensitive, confidential, and business-critical information.
AI Quality & Evaluation
- Develop processes to test and evaluate AI outputs.
- Create evaluation criteria for accuracy, relevance, consistency, grounding, and reliability.
- Identify hallucinations and incorrect knowledge retrieval.
- Track recurring AI quality issues and coordinate improvements with technical teams.
- Support the creation of test datasets and evaluation benchmarks.
Cross-Functional Collaboration
- Work closely with AI/ML engineers, developers, SEO/content teams, operations, and business stakeholders.
- Translate business requirements into AI knowledge and governance requirements.
- Document AI processes so that non-technical teams can understand and use them.
- Help establish standardized AI workflows across departments.
Required Skills
- 3–5 years of experience in AI, AI architecture, knowledge management, data architecture, ML, or a closely related field.
- Strong understanding of Generative AI and LLMs.
- Practical understanding of RAG architecture.
- Knowledge of:
- LLMs
- Embeddings
- Vector databases
- Semantic search
- Knowledge graphs
- Prompt engineering
- AI evaluation
- Data governance
- Understanding of vector databases such as Pinecone, Weaviate, Qdrant, Milvus, Chroma, or equivalent.
- Familiarity with LLM platforms/APIs such as OpenAI, Anthropic, Gemini, or similar.
- Strong documentation and analytical skills.
- Ability to understand technical architecture and communicate it clearly to business teams.
- Experience working with structured and unstructured data.
Good to Have
- Experience building or managing production RAG systems.
- Experience with knowledge graphs or ontology design.
- Familiarity with AI governance frameworks and responsible AI practices.
- Experience with LangChain, LlamaIndex, Haystack, or similar frameworks.
- Basic Python/API knowledge.
- Experience evaluating LLM applications using automated or human evaluation methods.
- Experience working with enterprise knowledge management systems.
What We Are Looking For
We are looking for someone who can actually understand and work with AI knowledge systems, not someone whose experience is limited to using ChatGPT or writing prompts.
The candidate should be able to explain:
- How a RAG system works end-to-end.
- How to improve poor retrieval results.
- How embeddings and vector search work.
- How to reduce LLM hallucinations.
- How to determine whether information is trustworthy enough for an AI system.
- How AI knowledge should be governed and maintained over time.
- How to evaluate whether an AI application is producing reliable answers.
Candidate Profile
Strong fit if you are:
- Hands-on with LLM/RAG systems.
- Comfortable discussing AI architecture.
- Strong in knowledge organization and data quality.
- Interested in AI governance and reliability.
- Able to bridge technical and business requirements.
Click on Apply to know more.
