Human Fluency Lab × Praxium × NTU Care Serve Learn
Designing a human skills program for 400 NTU students preparing to serve vulnerable communities.
What happens when you need to prepare up to 400 university students to work with young people, including beneficiaries with special needs and trauma histories, and you have exactly one training day to do it?
You cannot rely on good intentions. You need a program students will actually engage with, that inexperienced facilitators can deliver with confidence, and that stays safe when it moves out of the classroom and into real relationships with vulnerable people. That was the design challenge behind The Future-Proof Human, developed by Human Fluency Lab for NTU's Care Serve Learn program, in partnership with Praxium.
At this stage, the story is about the build: how the program was designed, tested, challenged and refined before its first live delivery.
The challenge
NTU's Care Serve Learn program places undergraduate students into 40-hour service projects with young people, some from schools and some from homes, including young people with special needs and trauma histories. Up to 400 students would go through a single training day before starting their service year.
There was an immediate tension built into that. The training was pass or fail, and many of the students in the room would not have chosen this over their other commitments. At the same time, the young people they would go on to serve could be vulnerable, and their needs could easily extend past what an undergraduate facilitator was equipped to handle. So the training had to solve two problems at once: how do you make students want to engage, and how do you make sure what they learn can be used responsibly by people who are not trained counsellors or social workers?
Human Fluency Lab was brought in to design the training, develop the facilitator materials, and run the train-the-trainer program for the student facilitators. Praxium coordinated with NTU, recruited and managed facilitators, and developed the Emotional Bank Account module within the same overall architecture.
The build, step by step
Psychological framing, terms like individuation and emotional regulation, a self-development lens on the material.
An NTU student and an NUS student both rejected the framing, for different reasons. Neither saw a reason to engage.
Values became a personal decision filter. Relationships became a relationship maintenance system. Dreams became a direction statement with a date.
A university community service champion reviewed a live demo and surfaced a different risk: centring the volunteer's growth over the beneficiary's experience.
Facilitator guides, scripts, digital boards, printable booklets, trainer materials and slide decks, each in youth, PWD and senior versions.
Safety, metaphor, cross-document consistency, accessibility and representation reviews each found something a normal content pass would have missed.
Student facilitators experienced the full system as participants before practising every technique themselves.
Up to 400 NTU students go through the program, then carry it into a year of real service.
The program was first drafted around self-awareness language. Before that framing went anywhere near 400 students, it was tested through two structured interviews.
“Useless.”
Career-focused finance undergraduate, on the original self-awareness framingThe other interviewee, a well-connected NTU student active in her student union, was more open to the content but had a different objection: sitting through slides would lose her. She wanted the material active and hands-on, and she wanted to understand the tangible use before the activity, not after.
Two very different students, pointing at the same design problem. The program was asking students to engage with concepts before giving them a reason to care. So the psychological vocabulary came out, and the material was reframed around outcomes students could immediately place in the context of their own lives, careers and relationships.
Students identify a lived value and an aspirational one, and build an action step from the evidence.
Human Fluency LabStudents map deposits and withdrawals in key relationships, then commit to one deposit action.
Developed by PraxiumAn image carousel exercise moves students from gut response to a dated commitment.
Human Fluency LabRather than asking students to complete all three, each session lets them choose two. A four-hour training day could easily turn into an exercise in cramming too much content into too little time, and giving students a choice let each module keep enough depth to mean something, instead of turning the day into a content marathon.
Once a working version of the Values Module existed, it was tested again with a university community service champion, whose volunteering history ran considerably longer than either of the first two interviewees.
Her perspective surfaced a different set of issues. She pointed out that volunteering programs can leave volunteers unclear about why their presence even matters, hand them safety instructions without ever teaching them how to build rapport, and sometimes close with a self-focused takeaway along the lines of “I'm grateful for what I have.”
A service program can, without meaning to, turn the beneficiary into a vehicle for the volunteer's own personal growth.
Design risk named by the community service champion's reviewHer response to the demo itself was positive: approachable, gamified without being try-hard, free of the preachy self-development tone she'd encountered elsewhere. She also picked up on something specific to the facilitator experience, that pairing a word-for-word script with a stated rationale for each round meant the facilitator wasn't just following instructions, they could understand what each part of the session was actually trying to do.
Her feedback became concrete changes: rapport-building guidance alongside the existing safety guidance, an explicit statement of why the facilitator's presence matters, a closing debrief reviewed to keep it from centring the volunteer over the beneficiary, the year-long nature of the commitment made explicit early, and her own suggestion, that facilitators experience the program as participants before delivering it, built directly into the 18 August train-the-trainer session.
The result is a program tested from three angles before its first live delivery: the student who has to sit through it, the volunteer who has to serve, and the facilitator who has to make the experience work in the room.
Designed for the facilitator, not just the participant
The facilitators are undergraduate students too, so the structure carries the judgment they haven't built yet.
Opening activity, same for every group
Deeper activity, basic version
Deeper activity, advanced version
Closing reflection: written declaration + one action
A group that isn't ready to go deeper isn't forced there. A group that's ready for more isn't held back. Either way, the facilitator has a structure to lean on rather than having to improvise the call in the moment.
Safety, designed in, not bolted on
That is what the three-tier response system is for.
End the activity and get appropriate support involved. No probing.
Dial back the intensity, give space, proceed carefully, check in privately.
Watch for disengagement or a shift in participation, without escalating.
Non-negotiable: an adult, teacher, support worker or appropriate staff member is present whenever the program is delivered to actual beneficiaries. The student facilitator is there to facilitate the program. They are not there to become a counsellor.
The program was never a single slide deck. It's an ecosystem: facilitator guides, word-for-word facilitator scripts, digital group boards, printable participant booklets, trainer guides, trainer scripts, and slide decks. Because the eventual beneficiaries aren't one homogeneous group, all of it exists in three versions, for youth, for persons with disabilities, and for seniors.
The 5 September training runs on the youth version. NTU student facilitators experience the program as participants first, then practise facilitating it themselves. The 18 August train-the-trainer session introduced the student facilitators to the full system, with every technique demonstrated before trainees practised the matching activity.
The build was deliberately uncomfortable
A program at this scale can't be validated by asking whether the slides look good. Five reviews turned up problems an ordinary content pass would have missed.
Found: a question asking who participants could tell about their reflection had no opt out.
Fixed: the question was removed, and the same opt-out standard applied across all three modules.
Found: the deposits and withdrawals language raised a stakeholder concern with NTU.
Fixed: rewritten around relationship building and breaking, round names and mechanics included, in two passes.
Found: a shared action removed from the digital board still appeared live in three other documents.
Fixed: five documents rewritten against the current board, not just the one where the gap was noticed.
Found: the seniors materials had smaller print than the PWD materials.
Fixed: font size, line spacing and contrast corrected against WCAG 2.1 AA, plus a click-to-enlarge on every board.
Found: 38 images per population audited against an ethnicity mix target found more than a quarter of the disability images defaulting to wheelchair users, one near-duplicate image, and skin-tone-only representation with no other visual cue.
Fixed: each gap given a specific corrective image. The goal was never to look diverse, it was for participants to actually see themselves in the material.
The most interesting outcome of this design phase isn't the number of activities built or the volume of materials produced, it's the philosophy underneath them.
The content had to answer the question students were already asking, which is why does this matter to me.
If an undergraduate student has to facilitate the activity, the program can't depend on expert judgment the facilitator doesn't yet have.
The purpose of service isn't to manufacture a meaningful experience for the volunteer, it's to help facilitators build relationships that are actually useful and respectful to the people they're there to serve.
Safety, accessibility, representation and consistency weren't final checks bolted on after the program was "finished," they were how the weaknesses in the design itself got found.
Where the project stands, 27 August 2026
The next test is the one that actually matters. On 5 September 2026, up to 400 NTU students will go through the program before starting their 40-hour service projects with real beneficiaries. That's the first real chance to see whether the design works off the page, whether student facilitators can deliver it as intended, and whether the program builds the bridge it was meant to build between self-awareness and service. The year that follows will be the harder measure still, what happens once these ideas leave the training room and enter real relationships.
This case study will be updated twice, once after the first live delivery, and again once the service projects conclude.
The real measure of a human skills program was never whether the training works for four hours. It's whether something from those four hours is still showing up long after the room has emptied.
For organisations navigating AI
The four-part structure behind this build, diagnostic, program design, facilitator training, embedding, is the same one Human Fluency runs for organisations.
© Human Fluency Lab 2026. All rights reserved. Developed for the NTU Care Serve Learn program in collaboration with Praxium.