Shape taxonomy design and metadata modelling for Genesys Cloud AI knowledge retrieval at Miratech. Improve how agents, users, and AI find content.
As an Information Scientist, you will design information architecture for enterprise knowledge environments. The work covers taxonomy, metadata, content structure, and search optimization. You will translate analytical findings into practical recommendations supporting customer service agents, self-service users, and AI-powered solutions.
Key Responsibilities:
- Analyze knowledge bases for structural, metadata, and searchability gaps
- Assess 100-200 pilot articles for quality and tagging practices
- Design taxonomies, controlled vocabularies, synonyms, and metadata models
- Define content structure and search optimization requirements
- Evaluate AI retrieval quality during proofs of concept
Requirements:
- 5+ years in information science, taxonomy, metadata, or enterprise search
- Experience designing enterprise content taxonomies and metadata models
- Practical experience analyzing search behavior and improving findability
- Experience preparing content for AI-powered search or virtual agents
- Eligibility to obtain and maintain Canadian Government security clearance
Apply information architecture, taxonomy, and metadata expertise to strengthen Genesys Cloud AI knowledge retrieval with Miratech.
Information Architect Taxonomy Specialist in Calgary employer: Miratech
Miratech is an exceptional employer that fosters a culture of Relentless Performance, ensuring a 99% project success rate and over 30% year-over-year growth. With a commitment to employee development through mentorship and certification programs, as well as a welcoming multicultural environment, team members can thrive while working on impactful projects for top global clients. The flexibility of a Work From Anywhere culture, combined with comprehensive benefits and a focus on social sustainability, makes Miratech an attractive choice for those seeking meaningful and rewarding employment in the tech industry.
