Skip to main content
Training platform

OMNI University

The same training engine deployed for a brand that needed learning to feel like part of the product: upload video, build the assessment, assign by role, and track who passed.

Specialty coffee
Client
Training platform
Engagement
Custom Software
Service
Video first
Because you cannot teach technique in a slide deck
By role
Assignment driven by job title and area rather than a catalogue
Tracked
Pass and fail recorded per person, with failures flagged to the team

Walkthrough

OMNI University in use

The interface the team works in every day.

0:00 / 0:00
OMNI University — Training platform

The challenge

What we walked into

Teaching a craft is different from distributing documents. The team needed people to genuinely learn technique, not click through a compliance module, and generic course tools flattened the material into something that looked like everybody else and got treated accordingly.

Generic training gets treated as generic

People take training as seriously as it appears to have been taken. A module that looks like it came out of a template gets clicked through, because its own presentation signals that it is a box to tick.

That mattered more here than in most deployments, because the material was craft knowledge. Technique is learned by watching it done properly and then being checked, and neither of those survives being flattened into a slide deck with a quiz bolted on.

Video is the lesson, not an attachment

Authoring is built around uploading video, with questions attached to confirm the material landed. For teaching a physical craft that ordering is not a preference, it is the requirement, because the demonstration is the content.

Keeping authoring that simple also kept the library growing. The people who hold the expertise are not instructional designers, and any workflow that demanded more than record, upload, and add questions would have quietly stopped being used.

The right set for the right role

Courses are assigned from job title and area rather than presented as a catalogue. Somebody logs in and sees what applies to their job, which removes both the navigation problem and the possibility of the wrong training being assigned by whoever handled onboarding that week.

As roles change or courses are added, the assignment follows automatically. Nobody maintains a spreadsheet of who owes what.

Assessment that means something

Each course resolves to a pass or a fail against its questions, recorded per person. A failure notifies the team so it can be addressed, and the failed section stays locked until it is reopened rather than being immediately retakeable.

That lock is what keeps the result honest. Without it, an assessment becomes a memory test of which answer the system accepts, and the record it produces is worthless.

Administrators can see the whole picture: who was assigned what, who completed it, and how they did.

How we went at it

  • Designed the interface around the brand so training reads as part of the product
  • Built authoring around video upload, because the demonstration is the lesson
  • Assigned courses from job title and area rather than a browsable catalogue
  • Recorded pass and fail per person, with team notification on failure
  • Locked failed sections until reopened so results stay meaningful

What we handed over

  • Fully branded learning environment
  • Video course authoring with attached questions
  • Role and area based course assignment
  • Individual logins with completion and result tracking
  • Failure notification and retake locking
  • Admin visibility into assignment, completion, and results

What happened next

  • Training looked like the company rather than a compliance tool
  • Craft technique was taught by demonstration and then actually checked
  • Assignment stopped depending on who ran onboarding
  • The team could see who had genuinely passed, not just who had clicked through

Capabilities

  • Product design
  • Video learning platform
  • Role-based assignment
  • Assessment and tracking

Stack

  • Node.js
  • JavaScript
  • React
  • Vite
  • Redux
  • React Query
  • Tailwind CSS
  • MongoDB
  • Video streaming
  • NodeMailer
  • AWS for the API and video
  • Vercel for the frontend

Find out what should actually be built.

Start with an assessment. We walk your business end to end and show you where automation and AI pay off, ranked by what they are worth.