Enterprise-grade technical curriculum. Built in days, not quarters.
Imoya is an agentic AI courseware platform. Your experts bring the knowledge; Imoya's agents draft the lessons, labs, exams and video; your experts approve every stage; you publish under your own brand.
Licensed per organization: each customer gets its own tenant, authors and reviewers.
The platform
Your experts bring the knowledge and keep final say. Imoya's agents do the drafting at every stage, from source material to certification-grade courseware.
Agentic Simulation Engine
Specialized agents turn learning objectives into interactive labs and scenario environments, so learners practice the task instead of reading about it.
3D Spatial Explanations
3D world-understanding models render architectures, data flows, and hardware as animated, explorable explanations generated from the lesson script.
Dynamic AI Tutoring
A course-grounded tutor answers learner questions from approved content only and focuses practice where each learner struggles.
RAG-Grounded Curriculum Generation
Every lesson, slide, and script is retrieved from and cited to your source material: vendor docs, blueprints, and internal runbooks.
Agentic Multi-Agent Workflows
Planner, author, reviewer, and assessment agents work in parallel across a course, so a full curriculum moves through the pipeline in days.
Adaptive Assessment Engine
Timed practice exams and dynamic evaluations, mapped to each objective, that measure mastery rather than a raw score. In production today: it powers NavyDuck.
Neural Instructor Avatars
Lip-synced avatar video of your own instructors with neural narration, rendered with GPU inference.
Human-Anchored AI Pipelines
Your subject-matter experts approve each stage before anything ships, and full provenance traces every claim back to its source.
Multi-Tenant Delivery
Your own branded tenant, team collaboration, and private or commercial distribution under your organization's name.
Powered by GPU inference
Imoya's pipeline runs open-weight models on containerized GPU inference, with avatar video and narration rendered on GPUs.
Inference
- Containerized vLLM & SGLang serving
- Open-weight models
- NVIDIA L40S GPUs
- Red Hat OpenShift
Rendering
- Local GPU avatar lip-sync
- Programmatic video assembly
- Neural text-to-speech
Orchestration
- Autonomous agent framework
- Planner · author · reviewer · assessor
- Human approval gates
- Full-curriculum batch pipelines
Grounding
- RAG over customer source material
- Citation-level provenance
- Tenant-isolated content stores
- Learning-objective mapping
Design goals: predictable GPU-hour unit economics, and customer source material kept on dedicated infrastructure.
Proven in production: NavyDuck
NavyDuck is Imoya's own certification-prep brand and the first product built on this pipeline.
It offers 27,000+ original practice questions across 96 exams — Cisco, CompTIA, AWS, Microsoft, NVIDIA, IBM watsonx, GitHub, Databricks, and others — with timed mock exams and instant explanations.
Questions are original and written to cover each exam's objectives.
Engineered by technical instructors, for technical instructors
Decades of enterprise architecture, certification authorship, and classroom delivery, built into the pipeline.
John Swartz
CCIE #4426 since 1999, Sybex author, Cisco Certified Systems Instructor, and NVIDIA-certified in AI Infrastructure & Operations. Architect and lead engineer of the Imoya platform and its GPU pipeline. His doctoral research at GWU on reducing curriculum development cost with generative AI is the foundation of Imoya.
Todd Lammle
Author of the Sybex CCNA study guides and a Cisco instructor. Leads curriculum quality and the expert-validation standard every course passes through.
Built by people who have written the books, taught the classes, and passed the exams.
Built for enterprise training teams
Early access is opening to IT, cloud, and AI training providers. Tell us where to reach you.
Or email imoya@imoya.com