Case study

Center360: one view of people, projects and staffing needs

A Capgemini centre with more than 600 people and international projects that come and go. Every new one starts with the same question: do we have this team in-house, or do we hire? I designed the internal platform that answers it, in a product trio with the PO and tech lead.

Role
Product Designer
Timeline
6 months
Team
PO, PM, Tech Lead, Devs
Platform
B2E · Desktop web
Outcome
≈60% faster profile search

01 · Context

The centre, and the question every new project starts with

Capgemini's centre in Valencia has more than 600 people on international projects that come and go. Each new one asks the same thing: do we have this team in-house, or do we hire? Answering it means knowing who is free, who has the skills, who is about to leave and which moves between projects would free the right person. That knowledge lived in spreadsheets, emails and a different application for every task.

THE STARTING POINTWHAT YOU NEED TO KNOWTHE SOLUTION A new project needs a team Staff it from inside, hire, or move people Vacancies What is open, and why:promotion, departure,change of role Projects Who is on which team,and who manages it Time off Who can actually joinon the project dates Team spirit Who is at risk ofleaving, before they do Center360 All four, in one place
The moves HR and staffing managers make every day, and the four things they need to know to make them.

We worked as a product trio: product, design and technology deciding together from the start, so every proposal was checked early for business sense, feasibility and real usefulness. I also kept a direct line to the centre's director, managers and HR.

The product trio. Three lenses on every decision before anything was built.

Lean product canvas for Center360: problem, solution, value proposition, segments, channels, costs and revenue
The lean product canvas we filled in with the centre's director before drawing a single screen.
Strategic decision

Build it in-house instead of licensing a tool. Three reasons:

  1. Long-term cost: no licences, no external dependencies.
  2. Training: the team gained real experience in React and Java while building it.
  3. Product potential: other Capgemini centres have the same problem, so it could be offered to them.

02 · Problem

The problem and its cost

Key information about projects, teams and skills was spread across Excel, email and tools that did not talk to each other. Each problem had a cost.

Information scattered and outdated
Wasted time and errors
No visibility of who is working on what
Hard to plan and assign staff
Manual, unscalable processes
Slow, person-dependent processes
Disconnect between areas
Little collaboration
No visible internal opportunities
Demotivation

03 · Research

Interviews, shadowing and four personas

Interviews and shadowing sessions to see how projects, staff and internal communication were actually managed. From the findings, four personas kept the design anchored to real needs, goals and frustrations.

Human Resources

Owns the data about people, and spends the day reconciling it.

Admin

Goals

  • Assign people without endless back and forth.
  • Coordinate with managers without friction.
  • Spot capacity gaps before they become urgent.
  • One reliable source for all employee data.

Frustrations

  • Requests that arrive vague or without context.
  • No full view of who is available.
  • Email, spreadsheets and meetings as the system.
  • Reactive processes, no way to anticipate.

Manager

Needs the right team on a project this week, not next month.

Admin

Goals

  • Build the right team fast.
  • See availability and skills at a glance, in and beyond the team.
  • Request staff without bureaucracy.
  • Decide on assignments quickly.

Frustrations

  • Cross-referencing Excel, Slack and meetings.
  • No clear place to see available profiles.
  • Depending on HR for operational questions.
  • Slow, inconsistent processes.

C-level

Decides on capacity and performance, and needs the whole picture to do it.

Super admin

Goals

  • A global view of projects and internal capacity.
  • Strategic decisions on current data.
  • Spot bottlenecks and operational risks.
  • Use internal talent better, reduce rotation.

Frustrations

  • Manual, outdated reports.
  • Key information not available in real time.
  • No visibility of how efficiently talent is used.
  • Hard to tell whether people are well distributed.

Employee

Wants to know what is going on, and a way to move.

Regular user

Goals

  • See which projects are active in the company.
  • Find internal opportunities that match their interests.
  • Apply to other projects on their own.
  • Feel part of the organisation and understand the bigger picture.

Frustrations

  • Not knowing what other teams are doing.
  • No clear channel to join a project or change team.
  • Feeling disconnected, off the radar.
  • Little internal transparency.

04 · Goals

Goals for users and for the business

For users

  • Faster staff selection.
  • No duplicated data, fewer errors.
  • Organisational transparency.
  • A stronger sense of belonging.

For the business

  • Lower talent rotation.
  • Leaner internal processes.
  • Better use of available talent.
  • A stronger organisational culture.

05 · Architecture

Role-based architecture and permissions

Once the main functionalities were validated, the question was how each profile reaches them. We structured the system as a funnel: visibility and the ability to act widen with the profile, from employees who explore and apply, to C-level with a global view of capacity and performance.

Super admin C-level and management
Admin BU leads, people leads, HR, university, managers
Regular user Employees
Who sees and does what.
Sitemap annotated with permissions per role
Sitemap with permissions per role.

06 · Decisions

Wireframes, and how decisions were made

After the first focus groups, I designed wireframes so the conversation could be about something tangible. They confirmed interest in several proposals and surfaced needs specific to each role. Dozens of decisions came out of those sessions, big and small, and they all followed the same process: name the constraints, weigh the options, pick one and say why. Three examples of that process, drawn as decision trees:

Focus group conclusions with project managers
Project managers.
Focus group conclusions with C-level and business units
C-level and business units.

Where should the main menu live?

QUESTION OPTIONS WEIGHED DECISION Where does navigation live? New modules are already planned Top bar Dropped · runs out of room as modules grow Sidebar menu Chosen · scales with new sections Menu behind an icon Dropped · hides core navigation on desktop Sidebar Room to grow,always visible
  • ConstraintRoadmap of new modules
  • ConstraintDesktop-first, dense screens
Path takenRejected, with the reasonConstraint

How many columns does the vacancies table need?

The hardest table in the product. Left to the managers, it would have had fifteen columns. The work was deciding what is essential to act on a vacancy, what can live in the detail page, and how to keep related data together.

QUESTION OPTIONS WEIGHED DECISION Vacancies table Show what is needed to decide, no more Every column they asked for Dropped · horizontal scroll, nothing stands out Few columns, rest in the detail Dropped · one click per vacancy to compare Merge related data in one cell Chosen · eight columns, nothing lost Eight columns Role cell carriesstatus, name,language, stack
  • ConstraintManagers asked for ~15 columns
  • ConstraintScanned daily, on laptops

RESULT · MIXING RELATED DATA IN ONE CELL

  1. Status as a dot, not a column. Green: the vacancy is open.
  2. Role name, the one thing every manager reads first.
  3. Language requirement as a chip in its own colour: managers filter by it, and it would otherwise be lost among skills.
  4. Technologies as chips, because for developer roles they decide the fit.
Path takenRejected, with the reasonConstraint

Where do managers mark a prospective candidate?

QUESTION OPTIONS WEIGHED DECISION Marking a candidate Speed up the decision Inside the candidate detail Dropped · several clicks per decision Edit from the vacancy table Chosen · decide while comparing Email HR Dropped · untraceable, the old way Inline edit Prospective flagset from the table
  • ConstraintDecisions happen while comparing
  • ConstraintMust stay traceable
Path takenRejected, with the reasonConstraint

The inline edit came as the next step, once the table had settled. One more iteration: a more visual user profile with a graphic view of a person's project assignments.

07 · Outcome

The outcome

Vacancies overview
Vacancies. Required technologies and languages as chips; prospective candidates editable from the row.
Projects overview as a kanban by status: upcoming, ongoing, delivered
Projects, as a kanban by status.
Employee profile with the time off calendar
Employee profile.
Team spirit page
Team spirit.
Team calendar showing public holidays, approved and requested vacations and absences for every team member
Team calendar. Holidays, vacations and absences for the whole team, in one view.

08 · Impact

Impact

≈60%*less time for managers to find a profile
4 → 1parallel systems replaced by a single source of information
≈ 0clarification emails between areas after launch
Voluntaryadoption by employees, driven by transparency and internal visibility

Staff assignment errors dropped because information was current and traceable. Participation was high in the first weeks, with positive feedback on clarity and usefulness.

* Estimated by the managers who used it, comparing a typical profile search before (spreadsheets and email) and after, in Center360. Not a controlled measurement.

09 · Lessons

Lessons and next steps

  • Balance client and user needs. Products move the metrics that matter when both are served, not one.
  • Involve every profile from the start. It is what made the tool relevant at every level of the organisation.
  • Product trio. Business, feasibility and experience weighed in the same conversation.
  • Design for low availability. C-level and managers had little time; asynchronous validation, short interviews and concrete visual deliverables solved it.
Next steps spotted
  1. Suggest compatible profiles for a project automatically, from skills, availability and history.
  2. Performance tracking: merge the existing employee evaluation platform and use its data for monitoring.
  3. Real-time KPI dashboard for C-level with proactive alerts.