A centralized career platform for the University of Michigan School of Information — reorganizing fragmented job boards, mentorship, and events into one tool tailored to how UMSI students actually search.
How might we leverage and reorganize UMSI's resources and network to increase the success rate of finding internships and post-graduate opportunities — while decreasing the cognitive load for overwhelmed students?
"Resources such as job boards, career events, and mentorship programs are scattered across multiple platforms, creating a fragmented and inefficient system."
UMSI students don't lack resources — they lack a coherent way to find, sequence, and trust the ones already available to them. Handshake, the UMSI Slack, the Engaged Learning newsletter, individual office emails, Iris alumni search, peer Google Docs: each is useful in isolation, but together they form a maze that disproportionately punishes students with less social capital and less time.
Students reported actively using six or more platforms during a typical week of job searching.
Every interview participant described a constant background feeling of being behind on something.
Multiple participants said they'd only used UMSI's career resources once — during first-year orientation.
Only 2 of 8 had ever reached out to a UMSI alum, despite all of them naming mentorship as critical.
Anchor ran on a compressed 12-week academic timeline, so the team designed the process around three forcing functions: every research artifact had to inform a design decision, every design decision had to be testable, and every test had to feed back into a single shared source of truth.
We used a mixed-methods approach — qualitative interviews to surface the lived experience, and structured usability protocols to validate the design once it took shape.
Conversations with Engaged Learning, Career Development, and DEI offices to understand each surface students touch.
8 UMSI students across MSI tracks (MHI, Big Data, Agile Development, LAKES) screened for varied career stages and visa status. Conducted over Zoom, recorded with Loom.
Quote-level observations clustered into four primary themes that anchored the design phase: fragmentation, mentorship opacity, visa invisibility, cognitive overload.
Current-state map across five stages of the search, with emotion overlays and pain-point tagging.
Audit of Handshake, CareerLink, LinkedIn, Iris (alumni), and peer-school platforms — what each does and where it breaks for UMSI's audience.
Moderated tests with 8 students using a written protocol, two core flows, and a post-task SUS survey.
The affinity work consolidated into four themes that became the design team's shared vocabulary for the rest of the semester. Every screen, every interaction, every cut had to trace back to one of these.
Students described "knowing where everything lives" as a survival skill — and one that disproportionately advantages students with existing networks or institutional fluency.
Every student named mentorship as critical, but only two had ever messaged an alum. Blockers were rarely interest — they were knowing who to ask, how, and how much was reasonable to ask for.
International students did meaningful additional work to filter sponsoring employers — most of it manual, much of it duplicated across friends. None of the existing UMSI tools surfaced sponsorship status reliably.
Beyond fragmentation, students described a constant background hum of "am I missing something?" — and then avoided UMSI's tools altogether after early failed attempts to use them.
Three personas captured the most common UMSI career-search modes. They were the team's tie-breaker tool — every contested design decision had to be defensible against at least two of these three.
Build a professional network, stay organized, and clarify career pathways into UX.
Disjointed resources, generic mentorship opportunities, and uncertainty about her progress.
Cher's persona shaped Anchor's centralized dashboard and tailored mentorship features.
Secure visa-sponsored roles, connect with mentors, and manage limited time effectively.
Generic job boards, intimidating networking processes, and time constraints around OPT.
Dave inspired advanced job filters and mentorship features designed for international students.
Support students with confidence-building tools and improve the internship pipeline.
Existing internship tools and mentorship opportunities lack depth and continuity.
Jane's insights inspired industry-standard features and ongoing mentorship rather than one-off coffee chats.
We mapped the typical UMSI career search across five stages. Each stage surfaces what students actually do, the tools they reach for, the emotional state they're in, and the design opportunities Anchor took on.
The four themes and journey map pointed at one underlying truth: UMSI doesn't have a discovery problem, it has a coherence problem. The resources are already there. We re-stated the brief as a How Might We to keep the design conversations grounded in that distinction.
To keep an opinionated team aligned across four parallel workstreams, we agreed on three design principles up front. Every screen review used these as the rubric.
If Anchor surfaces a role, event, or alum, students should believe it's been vetted for UMSI fit. Quiet authority beats loud marketing.
Every "first action" — first message, first RSVP, first save — should take fewer than three taps and one minute.
Students are not casual users. Anchor optimizes for repeat, focused sessions — not engagement metrics or notification load.
Combine job boards, career events, and mentorship into one platform — eliminating tab-juggling across UMSI tools while preserving the office-by-office context where it matters.
Mentorship matching and filters for visa sponsorship, role type, and student-rated mentor helpfulness — making invisible information visible.
Reduce cognitive load with streamlined navigation, clearly defined pathways for each user action, and consistent visual language across every flow.
The IA reorganized UMSI's existing resources around three primary hubs — Feed, Events, and Jobs — with a mentorship layer that connects students to alumni through every surface, not as a standalone tab.


Before building screens, we mapped 11 end-to-end user flows across the four primary surfaces. Each flow traces a single intent — sign in, RSVP to a session, send a mentor request, cancel a registration, save a job — from entry to confirmation, with explicit error-prevention states for cancellation flows.
We worked in three deliberate stages, validating with peers between each. Holding off on visual polish until the third stage kept critique focused on flow and clarity, not aesthetics.
Pen-on-paper sketches per primary tab (Feed, Mentor, Jobs, Events). Layout decisions made fast, before any pixel got pushed.
Grayscale Figma frames testing IA, button hierarchy, and content density across all four tabs. Validated in the first round of usability testing.
Branded Anchor screens with the UMich navy/maize system, real strings, photography, and full interactive prototype wiring.
Anchor's visual language leans into Michigan's identity — navy authority, maize energy — paired with a quieter Mulish/Fraunces type system that lets the content lead.
Color-blindness-friendly contrast pairs informed every primary/secondary combination.


The Feed is a low-pressure hub for students to connect with peers and alumni, see referral posts, and share milestones as they progress through the year.
Designed for UMSI's culture, not retrofitted from a generic professional network — every component asks the question, "Would a UMSI student feel comfortable posting this?"
The Events surface unifies the Engaged Learning, Career Development, and DEI office calendars into a single, filterable view — with a built-in scheduler so students don't have to leave Anchor to RSVP or track an upcoming session.
Tagging events by office keeps each program's identity intact while removing the cost of knowing which office hosts what.
The Jobs page surfaces internships and full-time roles relevant to UMSI students — Product Designer, Data Analyst, Design Engineer, UX Researcher. The referrals view highlights what a hiring manager is actually looking for: required major, project experience, and coursework, so students can self-screen before they apply.
Michigan's alumni base is enormous but historically hard to navigate. The Mentorship flow lets students filter alumni by career goals, read peer-rated helpfulness, send a request using a starter template, and book a meeting directly — turning a passive network into an active one.
Three flows did most of the work to make Anchor feel different from CareerLink or Handshake. Each came directly out of an interview insight and was validated through usability testing.
A formal mentorship program rather than a one-off coffee chat. Alumni profiles include customizable expectations for mentees — relevant coursework, projects, or skills — so students can align their goals with industry demand before they even reach out.
Group mentoring sessions, real-time chat, and asynchronous message boards — designed in response to feedback that the career search felt isolating. Inclusive of varying schedules and communication preferences.
A single surface for tracking deadlines, application statuses, and employer feedback — consolidating what CareerLink and Handshake forced students to do across separate logins. Timeline structures and reminders are tuned to internship cycles.
Testing surfaced two design decisions that we'd convinced ourselves were "obvious enough" — and weren't. Both got rewritten before the final stakeholder readout.
The home was a static "resources" page — a tab students described as opaque and difficult to navigate. Resource use dropped off after orientation.
Replaced with a social-feed format surfacing primary, secondary, and tertiary connections. Borrowed familiar patterns (Jakob's Law) to lift retention and daily active use.
Users disliked scrolling through long job lists with no way to narrow them down to UMSI-relevant roles.
Added a search bar and filters tuned for UMSI students — leveraging Hick's Law and selective attention to reduce cognitive load and improve click-through.
Conceptually clean — but users repeatedly missed the section and described referrals as "a thing I heard about but couldn't find."
Promoted referrals to their own surface with a "what they look for" structured field on every card — required major, project experience, coursework.
We tested with 8 UMSI students recruited across MSI tracks (MHI, Big Data, Agile Development, LAKES). Sessions ran 15–20 minutes over Zoom, recorded with Loom; one in-person session was also captured. Participants worked through tasks against low-fi Figma prototypes.
Filtering alumni, reviewing peer-rated helpfulness, and sending a request.
Locating the right office's events and RSVPing in-platform.
Navigating from a referral card to an application with the right context.
Maintaining a persistent shortlist and a profile that mentors could actually act on.
We tracked error rates, success rates, and time on task, then ran a short post-session interview to surface the why behind each metric.
Find a saved job posting from the Feed and book a mentorship session with an alum. Designed to stress both navigation and trust-building.
Students universally appreciated the centralized platform and reported that mentorship felt approachable for the first time.
Some users struggled with the referrals section and requested clearer terminology — addressed in the v2 iteration described above.
Two of the testimonials we received from UMSI students during testing and post-readout sharing:


The complete case-study prototype is embedded below — click through to test sign-up, Feed posting, event RSVPs, job applications, and mentor requests end-to-end. Error-prevention modals for cancellation flows are wired in.
Anchor was my first client-facing UX research engagement, and it taught me that the real deliverable wasn't the prototype — it was the shared model of why UMSI's career experience felt fractured. Once stakeholders saw the friction mapped end-to-end, the design decisions became almost self-evident.
I left this project more confident that good research design upstream lets the visual design downstream be quiet and obvious. The next time I take on a project this complex, I'll spend more time on the front end — building the journey map and design principles together with stakeholders, not for them — and less time defending screens.
UX Researcher & Designer · interview lead · journey map · Mentorship flow owner
Ciani Foy · Feyi Apampa · Niket Kamat Satoskar · Sabrina May
Figma · FigJam · Notion · Otter · Google Workspace