Redesigning "Learning, Smarter," the 0→1 product rebuild behind Assistiv LMS
FTF Platforms · Enterprise B2B · 2026
Responsibilities
Led research, IA, and design across all three user roles. Collaborated with CEO, engineers, and motion designer.
Project Timeline
Jan – May 2026
Made for
Assistiv - an AI-powered LMS serving compliance-focused B2B organizations: workforce reentry, mandated healthcare training, corporate onboarding.
What's Assistiv?
Assistiv is an AI-powered learning management system built for organizations where compliance isn't optional, workforce reentry programs, mandated healthcare training, and corporate onboarding. Their tagline: "Learning, Smarter." Admins manage learners. Instructors run courses. Students learn through AI-assisted study spaces. The platform is fully white-labeled, and each client organization gets the product under its own brand.
When I joined FTF Platforms, all three user types were navigating the same interface. None of it was built for any of them specifically.
Final Designs
Title
Why did we need a redesign?
As Assistiv expanded beyond one user type, the original navigation built for a single role started quietly breaking for everyone else.
Before - One for all Roles
Institute Overview
Users and Roles
Courses and Groups
Grades
AI Usage
Self Study
Spaces
System Setup
Admin, Instructor, and Student navigated the same bar. Nobody owned any of it.
After - Three purpose-built systems
Admin: Institute Overview
Admin: Users and Roles
Admin: AI Usage
Admin: System Setup
Instructor: Courses and Groups
Instructor: Grades
Learner: Self Study
Learner: Spaces
Learner: Courses
Each role sees only what their job require. A role switcher handles dual - permission users.
So I Rebuilt
Assistiv
A role-separated, white-labeled experience where Admins oversee, Instructors teach, and Students learn each in an interface built for their actual job, under each client's own brand.
Role-based navigation
Custom role builder
White Labeling
Instructor home
AI study Spaces
Admin Dashboard
Permission Marix
AI usage montoring
Onboarding flow
Research
Four organizations. One unfiltered conversation.
Before opening Figma, I led a structured partner research session with four live Assistiv clients.
Project Restart, Pelican Bayou, Calibre Health, and Alliance of Elite Leadership. I framed it as a UX research conversation, not a demo, obtained recording consent, and asked structured questions about workflows and what people needed the moment they logged in.
Issue 1
Issue 2
Issue 3
"We need to see who's falling behind and talk about it in funder meetings — not after digging through tables."
— Brittney, Pelican Bayou & Calibre Health
"Do I create a course or a group first?" "We need front-to-back tutorials. Even examples of when to use groups versus courses."
— Asked by every new admin, unprompted
"She decided to abandon the LMS due to the inability to find necessary features."
— Internal feedback, February 2026
#01
Admins needed signals. They were getting spreadsheets.
Partners logged in, wanting to answer "who do I act on today?" met with undifferentiated tables. No severity, no risk grouping, no "start here."
Raised by 4/4 organizations in the partner session
No way to surface at-risk learners without manual table scans
Role clarity had to come before the interface
Research confirmed the architecture was the problem, not the UI. Two weeks on Miro before opening Figma mapping permissions, ownership, and what each role's home screen needed to answer the moment they logged in.
Key Strategic Decision
Separate the navigation systems into three role-specific experiences on one platform rather than filtering items within a shared nav. A role switcher handles dual-permission users. The goal: each person feels like the product was built for their job, not modified to accommodate it.
Mid-project pivot - Enterprise B2B focus
Early designs skewed toward scholastic/academic metrics, detailing GPAs, academic timelines, and long-term cohorts. A Feb 2026 design review surfaced a critical misalignment: Assistiv's primary clients are B2B organizations (workforce reentry, compliance training, corporate onboarding), not academic institutions. The design focus was explicitly realigned to B2B-agnostic metrics, course completion, engagement, risk with institutional academic depth scoped for later. This single decision affected every data point shown on every screen.
Central thesis
Three roles. Three interfaces. One platform that finally works.
Mixing Admin, Instructor, and Student tasks in a single nav wasn't neutral; it made each person's experience measurably worse.
01
Surface the most urgent action first
Every home view answers "what do I need to do right now?" before anything else.
02
Make the AI visible at the right moment
Assistiv's differentiating feature needed to appear where work starts not buried in menus.
03
Close the intervention loop
Spotting a problem and acting on it should require zero navigation away from the screen.
Final Designs
Admin Dashboard
The first question on login is "Who needs my attention?", not How are we doing?"
Institute overview KPIs - Students requiring attention - Quick actions - course completion & Risk monitoring table. Three question. In that order. Every time.
1
2
3
5
4
1
Intervention surface not a report
Students requiring attention, groups by type, so admins act on patterns. One click reviews all affected students for that risk
2
Severity is visual first
High and medium tags make urgency immediate, so admins don't read descriptions to prioritize. They can scan the color and label in one pass
3
Shared courses surface automatically
"Common in CS301 or MAT201" means one intervention reaches many students. The dashboard cross references so the admin doesn't have to.
4
Givens deliberately scoped to version zero
Single events only, no reoccurring. Authentication + external sync would have blocked shipping
5
Table not chart b by design
Bar graph communicate trends. Tables communicate decisions
What I tried first?
Initial design used bar graphs. Early feedback flagged cognitive load. Switched to a color-coded risk table, cognitive load dropped significantly.
Studebt Profile + Reminder
Spotting a problem and acting on it should be on emotion
⚠️
Warning banner appears on profile
📧
Admin clicks Send Reminder
AI drafts the message
*or admin writes their own
📤
Sent via Outlook
📋
Logged in learner activity history
1
2
3
5
4
1
Intervention surface not a report
Students requiring attention, groups by type, so admins act on patterns. One click reviews all affected students for that risk
2
Severity is visual first
High and medium tags make urgency immediate, so admins don't read descriptions to prioritize. They can scan the color and label in one pass
3
Shared courses surface automatically
"Common in CS301 or MAT201" means one intervention reaches many students. The dashboard cross references so the admin doesn't have to.
4
Givens deliberately scoped to version zero
Single events only, no reoccurring. Authentication + external sync would have blocked shipping
5
Table not chart b by design
Bar graph communicate trends. Tables communicate decisions
Bar graphs communicate scale. tables communicate decisions. Admins weren't trying to understand a trend, they were deciding which course to act on next