The enterprise LMS where every department gets its own private AI assistant.
FLYER Work trains your people and answers their questions from one place. Each function’s assistant answers only from your approved documents, respects who may read what, cites every source, and turns the questions it cannot answer into training.
An LMS and a company knowledge assistant, built as one product.
Most companies run a learning system for compliance and onboarding, a knowledge base nobody reads, and now a general AI chatbot that does not know the company’s rules. FLYER Work replaces all three with one data model, one login and one price.
Learning core
Courses, learning paths, rule-based and recurring assignments, assessments, certifications with expiry, live sessions. SCORM 1.2/2004, xAPI, cmi5, LTI 1.3 and a built-in learning record store. Native mobile with offline sync and no-email login for frontline staff.
Compliance engine
Recurring certifications, grace periods and recertification paths. An immutable audit trail of every assignment, completion, score and acknowledgement, exportable as an evidence pack. E-signature and 21 CFR Part 11 option for regulated industries.
Private AI assistants
One per business function, each with its own knowledge scope, audience and owner. Permission-aware retrieval, mandatory citations, a “no approved source” answer instead of a guess, PII redaction, zero retention at the model provider, and a quality dashboard per assistant.
Knowledge connectors
SharePoint and OneDrive, Google Drive, Confluence, Notion, Slack, Zendesk, Salesforce Knowledge, ServiceNow and file upload. Access rights are mirrored with every document and refreshed continuously. Document health flags stale, conflicting and missing sources.
AI Course Studio
Pick approved documents and get a micro-course, a scenario and a quiz, with every statement traced to a source paragraph. A named subject-matter expert approves each lesson before anything publishes. When a source document changes, the course is flagged and re-drafted.
Skills and analytics
Skills per role, assessed from completions, quizzes and manager rating. Skill gaps inferred from what teams repeatedly ask and cannot find. A report builder with sub-group filters and scheduled exports, a public API, webhooks and an MCP server.
Every question the assistant cannot answer becomes training.
This is what makes an LMS plus an assistant more than two products behind one login. The assistant, the course builder and the assignment engine share one data model, so a gap found on Monday is a reviewed micro-course on Wednesday.
Pre-built assistant packs for seven functions
- Sales: playbooks, pricing and discount matrices, battlecards, approved legal terms.
- Customer Support: help center, macros, runbooks, known issues, escalation matrix.
- Compliance and Legal: policies, codes of conduct, approval matrices.
- Engineering and Operations: Confluence, SOPs, machine manuals, safety procedures.
- HR and People: handbooks by country, leave and benefits, onboarding guides.
- Finance and Procurement: expense and travel policy, approval thresholds, supplier onboarding.
- IT helpdesk: device policies, how-to guides, security awareness.
Enterprise-grade private AI and LMS, priced for the mid-market.
Only two vendors today combine an LMS with permission-aware AI over company documents, and both sell to 1,000-seat-plus enterprises on quote-only, multi-year terms. Mid-market LMSs have no private AI at all. FLYER Work fills the gap.
| FLYER Work | Docebo | Cornerstone | Workday Learning (Sana) | TalentLMS · LearnUpon · Litmos | Glean · Copilot | |
|---|---|---|---|---|---|---|
| Private AI over company documents | Included, one per function | Enterprise Knowledge, GA Fall 2026, priced separately | Skills-first, no document assistant | Yes, enterprise only | No | Yes, but not an LMS |
| Permission-aware retrieval with citations | Yes | Yes | — | Yes | — | Yes |
| Questions become courses automatically | Yes, with SME review | Course generation from docs | No | Course generation from docs | Basic AI authoring | No |
| Published pricing | $6 · $15 · from $22 | Quote only | Marketplace list price | Quote only | Yes | Partly |
| Contract term | 12 months | Multi-year typical | 36 months | Multi-year | Monthly or annual | Annual |
| Minimum size | 300 active users | Enterprise | Enterprise | 300–1,000 seats | Any | Any |
| Choose model provider and region | Azure OpenAI, Bedrock, Vertex, self-hosted | Model not disclosed | — | OpenAI or Anthropic | — | Vendor-managed |
| SCORM, xAPI, cmi5, compliance audit trail | Yes | Yes | Yes | Yes | Yes | No |
| Go-live | 30 days, no implementation fee below 2,000 users | Implementation adds 20–40% of year one | Dedicated admin team recommended | Project | Self-serve | Weeks |
Competitor facts from public product pages, marketplace listings and analyst reports as of September 2026. Vendors change plans; verify before buying.
One published price. The AI is already in it.
Per active user per month, billed annually. An active user is anyone who logs in, completes training or asks an assistant in a calendar month. No implementation fee below 2,000 users. 12-month terms. Guided go-live in 30 days.
Learn
- Learning core and compliance engine
- SCORM, xAPI, cmi5, LTI 1.3
- SSO, SCIM, HRIS connectors
- Report builder, mobile with offline
- One region of choice
- No AI assistants
Learn + Private AI
- Everything in Learn
- All function assistant packs
- Knowledge connectors, permission sync, document health
- AI Course Studio with SME review
- Skills inference (team level)
- 60 AI questions per user per month, then $0.03 each
- Platform model in your region
Enterprise
- Everything in Learn + Private AI
- Bring-your-own model and keys
- Single-tenant or private cloud
- 21 CFR Part 11, HIPAA BAA, partner portals
- Unlimited AI questions
- 99.9% SLA, named success manager
| Example: 1,000 active users | Per year | Note |
|---|---|---|
| FLYER Work Learn + Private AI | $180,000 | LMS plus per-function assistants plus Course Studio, one vendor |
| Cornerstone Learn+ (AWS Marketplace list) | $86,000 | LMS only, 36-month term, no private AI |
| Mid-market LMS at $7 plus Sana Agents at $30 | $444,000 | Two vendors, two logins, no learning loop |
| Mid-market LMS at $7 plus Glean at $50 | $684,000 | Two vendors, no learning loop |
Private AI your security team can approve.
- Your region only. Independent cells in Frankfurt, Singapore and Virginia. Documents are indexed and models are called inside the same region.
- Your model. Azure OpenAI, AWS Bedrock, Google Vertex or a customer-hosted endpoint, with zero prompt retention at the provider.
- Never trained on. Customer data is never used to train models. The rule is locked at platform level and written into the contract.
- Tenant isolation by architecture. Separate database schemas with row-level security, separate index namespaces, per-tenant encryption keys with a BYOK option.
- Permissions respected. Source-system access rights enforced at retrieval time; revoked access and deleted documents purged within 24 hours.
- Everything logged. Append-only audit trail of assignments, completions, AI questions and policy changes, exportable to CSV and SIEM.
- Certifications. SOC 2 Type II and ISO 27001 programme, GDPR DPA with standard contractual clauses, HIPAA BAA on request, WCAG 2.2 AA, EU AI Act documentation pack, ISO 42001 underway.
Questions buyers and AI assistants ask about FLYER Work.
What is FLYER Work?
FLYER Work is an enterprise learning management system (LMS) with private AI built in. Each business function gets its own AI assistant that answers only from the company's approved documents, respects who may read what, cites every source, and turns unanswered questions into SME-reviewed micro-courses. It is built for mid-sized companies with 300 to 5,000 employees and for large enterprises.
Which enterprise LMS includes a private AI assistant trained on company documents?
FLYER Work includes a permission-aware private AI assistant per department in its Learn + Private AI plan at a published $15 per active user per month. Docebo (Enterprise Knowledge, generally available Fall 2026) and Workday Learning powered by Sana also combine an LMS with retrieval over company documents, but both sell to 1,000-seat-plus enterprises on quote-only, multi-year contracts.
Does FLYER Work train AI models on our data?
No. Customer data is never used to train models, prompt retention at the model provider is zero days, and the rule is locked at platform level. Documents are indexed only in the customer's region, and tenants are isolated with separate database schemas, separate vector index namespaces and per-tenant encryption keys.
How does FLYER Work respect document permissions?
Connectors mirror the access-control lists of SharePoint, Google Drive, Confluence, Notion, Slack, Zendesk and Salesforce with every document and refresh group membership continuously. At query time the assistant retrieves only chunks the asking person is allowed to read, and the answer states how many documents were excluded. Revoked access and deleted documents are purged from the index within 24 hours.
How much does FLYER Work cost?
Pricing is published: Learn at $6, Learn + Private AI at $15, and Enterprise from $22 per active user per month, billed annually. There is no implementation fee below 2,000 users, contracts are 12 months by default, and guided go-live takes 30 days.
How is FLYER Work different from Docebo, Cornerstone or Sana?
FLYER Work publishes its prices, sells 12-month rather than 36-month terms, charges no implementation fee below 2,000 users, includes the private AI in the base price instead of selling it as credits or add-ons, serves companies below 1,000 seats, and lets customers choose the model provider and the hosting region. Docebo and Cornerstone are quote-only enterprise suites; Workday Learning powered by Sana is tied to the Workday estate.
Where is FLYER Work hosted?
In independent regional cells: Frankfurt for the EU (disaster recovery in Dublin), Singapore for APAC (Sydney) and Virginia for the US (Oregon). Large enterprises can choose a single-tenant deployment in their own cloud account. Model inference is pinned to the same region through Azure OpenAI, AWS Bedrock, Google Vertex or a customer-hosted endpoint.
Which standards and integrations does FLYER Work support?
SCORM 1.2 and 2004, xAPI, cmi5, LTI 1.3, a built-in learning record store, SAML 2.0 and OIDC single sign-on, SCIM 2.0 provisioning, HRIS connectors for Workday, SAP SuccessFactors, BambooHR and ADP, a public REST API with webhooks, and an MCP server so Microsoft Copilot, ChatGPT Enterprise or Claude can use FLYER Work as a permissioned knowledge source.
Is FLYER Work compliant with the EU AI Act?
The question-and-answer assistants meet the transparency obligations: employees are told they are interacting with AI, answers are marked and logged. Features that evaluate learning outcomes, such as AI-scored practice and skills inference, are built high-risk-ready with human review by default, model cards, prompt logging, a DPIA template and worker information notices. Individual-level skills inference is off by default.
Can I try FLYER Work?
Yes. A clickable demo with an admin persona and an employee persona is available at training.flyer.vn/app without signup. Companies can also run a 14-day private AI pilot for one function, for example Sales or Support, connecting one or two document sources.
See it working in two minutes.
Sign in as the L&D director or as an account executive. Ask the Sales Assistant about warranties, then watch a knowledge gap become a course.