
Skillsoft
Skillsoft is a global leader in corporate learning, providing digital training and education solutions to help businesses improve workforce productivity, reduce risk, and increase innovation.





.webp)
.webp)
A training management system in Malaysia or Singapore rarely gets abandoned. It gets quietly bypassed. Requests still get raised — over WhatsApp. Sessions still get scheduled — in a shared calendar. Attendance is still recorded — on a paper sheet that someone photographs and emails to HR on Friday. The platform sits there, licensed and largely unvisited, while the real process runs beside it.
That gap is a design outcome, not a motivation problem. Such a system gets used when it removes work from the people it touches, and gets routed around when it adds work so that someone else can produce a report. This guide is about building the first kind — the workflow, the evidence layer that HRD Corp and SkillsFuture claims depend on, and the specific design decisions that decide which way it goes.
Not software. A request-to-record workflow with one owner at each step: an employee or manager raises a training need, it is approved against budget and funding eligibility, a session is scheduled with a trainer and venue, attendance is captured at the session itself, the record lands against the employee's profile, and the cost is reconciled against the levy or grant it was meant to draw on.
Software makes that workflow faster. It cannot supply it. Organisations that pick a platform before defining those six steps end up automating whichever informal process happened to exist — which is why the spreadsheet survives the implementation.
Before anything else, one clarification that saves Singapore-based readers a wasted afternoon. In Singapore, "TMS" is a regulated term with a specific meaning, and a large share of what you will find searching for it is written for training providers selling funded courses — not for employers training their own staff. The distinction matters, and it is where this guide starts.
Read those four together and the shape of the build appears. Both markets tie money to evidence, and the evidence is generated at the moment a session happens — not in a report written afterwards. Any system that makes capture at that moment harder than a paper sheet will lose to the paper sheet, and the funding follows the paper.
Search the term and you get three unrelated products written for three different readers. Working out which you are is the first real decision, because two of the three are not built for an employer training its own staff.
Built for training companies that sell courses: course catalogues, public registrations, invoicing, payments, CRM and marketing. Most of the highest-ranking guides describe this.
Reader: a training provider running a businessA regulated category in Singapore. SkillsFuture Singapore requires funded training providers to run a TMS connected to SSG systems so enrolment and claims data flows automatically.
Reader: an SSG-funded registered training providerInternal requests and approvals, session scheduling, attendance, employee training records, cost tracking and funding claims. Usually a module inside an HR or learning platform.
Reader: you, if you employ the learnersThe third is what this guide builds — and the one nobody writes about for this region, which is why an HR manager in Kuala Lumpur or Jurong ends up reading about invoicing modules and course marketing funnels.
Because the SSG requirement is mandatory and well documented, employers occasionally conclude they are subject to it. They are almost never in scope. SkillsFuture Singapore's guidance applies the requirement to training providers offering SSG-approved funded courses, and explicitly excludes enterprises that do not offer training as a primary business but conduct in-house training for their own employees — noting those enterprises rely on their HR system instead.
In practice: sell SSG-funded courses to the public and you are in scope; train your own payroll and you are not, leaving you free to build what suits your operation. What still applies to you as a funded employer is the evidence side — including e-attendance via SingPass on classroom-facilitated and synchronous e-learning under SSG-funded courses.
| Question it answers | Training administration layer | Learning delivery layer |
|---|---|---|
| Core job | How training was requested, approved, scheduled, attended, paid for and evidenced | What content was delivered, completed and scored |
| Handles classroom sessions | Yes — trainers, venues, batches, calendars, capacity | Partially — usually as a calendar entry against a course |
| External providers | Central — vendor records, invoices, trainer credentials | Peripheral — external content is linked, not managed |
| Funding claims | Where the evidence pack is assembled | Contributes completion data only |
| Self-paced e-learning | Little involvement beyond enrolment and cost | Central |
| Fails when | Attendance and approvals live outside it | Content is stale or irrelevant to the role |
Most organisations here need both, because most training is blended: classroom safety and compliance sessions, externally delivered funded programmes, and self-paced digital content side by side. The mistake is assuming the delivery platform covers the administrative half. It usually does not, and that gap is where the spreadsheets breed. Our overview of what a learning management system covers sets out the delivery side in full, so you can see precisely where the boundary falls.
A practical test. Count last quarter's training events and mark each one as self-paced digital or as a scheduled session with a trainer. If more than about a third were scheduled sessions, your workload sits in the administrative layer, and choosing a platform on content-library size will solve the smaller half of your problem.
Before evaluating any product, write down how these six steps work today — including the parts that happen over chat. That is your specification. Judge everything a vendor shows you on whether it makes one of these steps shorter for the person who performs it.
Two deserve disproportionate attention. Step four sets your compliance position, because attendance is the evidence both funding bodies verify. Step two sets your recovery rate, because eligibility is decided before anyone books and cannot be applied retrospectively in either market.
The system is only as good as its slowest step, and the slowest step is almost always attendance capture at the door of a training room.
Notice what is absent: content. Choosing courses, building learning paths, and measuring skill growth matter enormously, but they belong to the delivery layer and run on a different cycle. Solving both in one project is the most common reason these implementations overrun. If content design is the more pressing problem, our guide to employee training programme types is the better starting point, and this build can follow it.
A training management system is used by three groups with completely different tolerances: employees who touch it a few times a year, managers who approve under time pressure, and administrators who live in it daily. Most implementations optimise for the third group, because they are the ones in the room when requirements are written. The first two then find a faster path, and the faster path is always WhatsApp.
| What people do instead | The real cause | The design fix |
|---|---|---|
| Request training over chat | Separate login nobody remembers; the form asks for a cost centre code the employee has never seen | Single sign-on from the HR system, pre-fill everything already known, three fields maximum |
| Approvals stall for weeks | Approval sits in an inbox behind a link, opening a desktop-only screen | Approve or decline from the notification itself, on a phone, with the funding flag shown inline |
| Paper attendance sheet | Digital capture needs each participant to log in, on venue wi-fi, before the session starts | Trainer-side roll call or a single session QR code; participant login is not a prerequisite for the record |
| Records typed up days later | The capture tool does not write to the record system, so someone re-keys it | One write, at the session; the record is created there, or it is not created |
| Managers keep a private tracker | No team-level view, so managers cannot see who is due what | A manager dashboard showing their people, expiries and pending approvals only |
| Requests are never raised at all | Nothing visible happened last time — no acknowledgement, no outcome, no explanation of a decline | Status visible to the requester at every step, including a reason when the answer is no |
| Frontline staff never appear | Access assumes a company email address and a laptop | Login by employee ID or phone number; supervisor-assisted enrolment for shared-device sites |
Beyond universal usability, three local realities decide adoption in Malaysia and Singapore, and each is easy to miss when configuring a platform designed elsewhere.
Language is not a settings toggle. A Malaysian manufacturing floor may run Bahasa Malaysia, Mandarin, Tamil and English across one shift; a Singapore site adds its own mix. Translating the interface while leaving the request form, notifications and evaluation questions in English produces a system that looks localised and reads as foreign. Notifications matter most — often the only part employees ever see.
Shift patterns beat calendars. Scheduling logic built around a nine-to-five office will collide with rotating shifts, plant shutdowns and two different public holiday calendars. Sessions that consistently land on a group's rest day teach that group to ignore training invitations, and that lesson is difficult to reverse.
Cross-border data handling is a live question. Both markets have a Personal Data Protection Act, and training records are employee personal data. Where records sit, who can view them across entities, and how long they are retained are build-time decisions, not audit-time questions — particularly for groups running one platform across both countries.
The test that predicts adoption. Before signing anything, have a real frontline employee — not an HR colleague — raise a request on their own phone while you time it silently. Then have their manager approve it. If the pair takes longer than three minutes, or either person asks you what something means, the rollout will underperform regardless of what the feature list promised.
None of these fixes needs a different platform — they are configuration and process decisions taken before launch. The same principles underpin engagement more broadly, covered in our guide to building for employee engagement, which covers in more depth.
This is what makes a regional build different from a generic one. In both markets public money attaches to training, released against documentation generated at specific moments in the workflow. Design around those moments and claims become a by-product; design around reporting and you will spend every quarter reconstructing evidence from calendars and email.
| Dimension | Malaysia — HRD Corp | Singapore — SkillsFuture |
|---|---|---|
| Nature of the money | A levy you have already paid — 1% of wages for employers with 10 or more Malaysian employees under the PSMB Act 2001 | A reduction on eligible course fees and defrayed costs, claimed after the fact |
| Economic behaviour | Sunk. An unused balance is forfeited value, so under-claiming is a direct loss | Foregone discount. Unclaimed support costs you the subsidy, not the principal |
| Approval timing | Grant must be applied for and approved in e-TRiS before training begins | PSG applications must precede any contract, deposit or payment — no retrospective approval |
| What must be registered | The individual course, under the HRD Corp Claimable Course scheme, delivered by a registered provider with a TTT-certified or accredited trainer | The course, listed in the SkillsFuture for Business Course Directory |
| Attendance evidence | T3 attendance forms; classroom sessions need daily signed records, online sessions need trainee details captured | E-attendance via SingPass for classroom-facilitated and synchronous e-learning on funded courses |
| Claim pack | Signed JD14 declaration, T3 attendance forms, evaluation forms, tax invoice, proof of payment | Invoices and payment evidence, submitted through the relevant scheme portal |
| Deadline | Six months from the training completion date | Scheme-dependent; SFEC claims are tied to the scheme's own validity window |
| Amendments | Attendance cannot be modified once approved | Corrections require resubmission through the portal |
The two schemes run in opposite directions, and a group with one blended budget across both countries will systematically get this wrong. Malaysia's levy is money already spent, so every unclaimed ringgit is forfeited value and the incentive is to plan training to be claimable. Singapore's support reduces a bill you would otherwise pay in full, so the incentive is to sequence approvals correctly and never let payment precede a Letter of Offer.
Malaysia penalises under-claiming. Singapore penalises out-of-sequence buying. One budget, two opposite failure modes.
The build consequence: eligibility must be a field on the request, not a note in a spreadsheet. When a manager approves a session, the system should already know whether it is claimable, under which scheme, and what evidence it must therefore produce.
Funding attaches to the course, not the platform. In Malaysia the programme must be registered under the HRD Corp Claimable Course scheme in e-TRiS by a registered provider; in Singapore only courses in the SkillsFuture for Business Course Directory attract support. Internally authored content therefore generally sits outside both schemes — not a bad idea, but it competes for unsubsidised budget while an external equivalent may not.
Verify before you rely on any of this. Both schemes change with budget cycles and circulars, and third-party summaries go stale quickly — the HRD Corp claim window in particular is misstated on several widely-read sites. Confirm current rules on hrdcorp.gov.my and skillsfuture.gov.sg before setting a policy, and note the dates you checked in your own documentation.
Getting this layer right settles the cost conversation too, since recovered funding is the fastest lever on net spend. Our guide to reducing training costs covers the levers that sit alongside it.
Systems fail on ownership more often than on features. When four people can each start a request and none owns attendance, records will be incomplete, however good the software is. Assign these five roles in writing before configuration begins.
Employee or manager. Raises the need. Owns nothing afterwards, which is exactly why the request step must be short, and the status must stay visible to them.
Line manager, plus a budget holder above a threshold. Owns the commitment of money and the timing decision that determines claimability.
The person who makes a session exist — trainer, venue, batch, invitations. Usually HR or L&D operations. The busiest role in the system.
Trainer or supervisor present in the room. Owns attendance capture at the moment it happens. Most organisations never name this role, which is why attendance data is weak.
Assembles and submits to e-TRiS or the relevant Singapore portal, and tracks disbursement. Often finance rather than HR — decide deliberately rather than by default.
Four record types, and the relationships between them, cover almost every requirement later described as a missing feature.
| Record | Must hold | Common omission that causes pain |
|---|---|---|
| Employee | ID, entity and country, department, cost centre, manager, employment type, preferred language | Employment type — contract and agency staff are excluded from the HR feed, then cannot be enrolled |
| Course | Title, provider, delivery mode, duration, cost, funding eligibility, validity period | Validity period — without it, expiring certifications are found during audits rather than before |
| Session | Date, trainer, venue or link, capacity, batch, country, linked course, cost | Country — a group running both markets cannot separate claim streams without it |
| Attendance | Participant, session, present or absent, capture timestamp, capture method, evidence file | Capture method — you cannot later prove whether a record was live or reconstructed |
The HR system, for people data. Joiners, movers and leavers should flow automatically. Without it someone maintains a parallel employee list; it drifts within a quarter, and training gets assigned to people who left. Plan deliberately for the exception: contract and agency staff who never appear in the HR feed but do appear in training rooms — build their enrolment path at the start, not as a workaround.
Finance, for cost and claim status. Training cost, invoice reference and claim outcome should reconcile against the session record. This is the connection that turns funding recovery from a quarterly investigation into a running number, and it is the one most often deferred to phase two and never built.
Everything else — calendars, video conferencing, single sign-on, e-signature — is valuable but replaceable. These two decide whether the system holds truth or an approximation of it. The mechanics of the first are in our guide to integrating learning platforms with an HRMS, which applies equally to the administrative layer.
Decide the source of truth once. If both the HR system and the training platform can create an employee record, they will disagree within a quarter, and nobody will know which is right. Name one system as authoritative for people data and one as authoritative for training records, write it down, and configure the other to read rather than write.
Most organisations shopping for a training management system reach for a new product when the answer sits unconfigured inside a platform they already pay for. Work through these three in order and stop at the first that fits — every extra platform adds a login, a data source and a reason to route around the process.
Most HR and learning platforms include a training administration module that is licensed and switched off. Configuring it gives you single sign-on free, one employee record instead of two, and no new vendor. For organisations up to a few thousand employees with a normal classroom-and-digital mix, it covers all six steps.
Right when: your session volume is moderate, your approvals are simple, and your people data already lives in one HR system Watch-out: confirm attendance capture and funding-eligibility tagging specifically — these are the two functions most often missing from bundled modulesJustified when instructor-led logistics dominate: many concurrent sessions, multiple venues, a roster of external trainers, resource conflicts, scheduling across sites. Bundled modules handle that complexity poorly, and dedicated tools exist precisely for it.
Right when: scheduling and resourcing consume more administrator time than everything else combined Watch-out: you now own an integration to the HR system and one to finance; budget for both before comparing licence costsDefensible only when a regulator, client contract or genuinely unusual operating model imposes a workflow no product supports. The licence saving is real; the maintenance cost is usually underestimated by a wide margin — patching, upgrades, handover, and the person who becomes the only one who understands it.
Right when: a written external requirement cannot be met any other way Watch-out: cost three years of maintenance and one developer departure into the comparison, not just the buildWhichever route you take, prove the highest-risk step before scale rather than after. That step is attendance capture in nearly every case.
Step six is the one organisations skip, and skipping it is why so many systems end up half-adopted: a parallel path that still works is a path people will use. Our guide to implementation strategy covers the change-management side, and choosing the right platform covers evaluation if configuration turns out not to be enough.
Login counts and course completions will not tell you whether a training management system has landed. Both can look healthy while the real process runs on paper beside the platform. What you need are process measures — indicators that reveal whether work is flowing through the system or around it.
| Indicator | What it reveals | Healthy direction |
|---|---|---|
| Requests raised in-system | Whether the front door is being used, or bypassed via chat and email | Rising toward near-total within two quarters |
| Median request-to-approval time | Whether approvals are a step or a bottleneck; long tails push people back to informal channels | Falling, with a shrinking tail |
| Attendance complete at session close | The single best proxy for whether capture design works in the room | Near-complete on the day, not on Friday |
| Records with capture method logged | Whether evidence is live or reconstructed — which is what an audit ultimately asks | Live capture dominant |
| Claimed against claimable | Funding recovery; the number that pays for the system in Malaysia | Gap closing quarter on quarter |
| Certifications expiring unrenewed | Whether the system is prompting the next cycle or waiting for an audit | Trending toward zero surprises |
An illustrative view for a regional group running both markets. Note that it reports process health rather than learning outcomes — those belong to the delivery layer and a different review cycle.
Both weak bars point at design gaps rather than effort. Fifty-seven percent claimed against claimable is money already paid to the levy and not recovered — the fix is tagging eligibility at the request step so it is never discovered late. Thirty-four percent contract-workforce completeness is the HR feed exclusion showing up exactly where it always does: people who work on your sites, attend your safety training, and exist in no system that governs it.
If attendance is captured on the day and eligibility is flagged before approval, most other numbers correct themselves. If either is broken, no dashboard will save you.
Once the process is stable, the conversation can move to whether the training itself is working — a different question with different methods, covered in our guides to measuring training effectiveness and measuring training ROI. Do not attempt both conversations in the same quarter.
Adoption is decided in the first quarter, mostly in the first three weeks. The failing pattern is a launch email, a training deck, and a review ninety days later where someone discovers half the sessions were never recorded. The working pattern treats the early weeks as active operations, not a monitoring period.
Attend sessions physically. Watch attendance capture happen with real participants and real connectivity, and fix friction the same week rather than logging it.
Approvals are where momentum dies quietly. Every request stuck for more than a few days teaches one manager and one employee that the system is slower than asking directly.
Take a single claimable session from request to submitted claim using only what the system holds. Every missing field will surface here, months before a deadline makes it expensive.
Only once the system demonstrably works, retire the parallel process. Doing this earlier punishes people for a system that is not ready; doing it never guarantees permanent half-adoption.
The most useful thing you can do in week two costs nothing. Ask three employees who raised a request what happened next. If any of them cannot tell you, the visibility problem is bigger than any feature gap, and it will suppress requests for the next year.
Sustained use also depends on the training being worth requesting — a system that runs perfectly while delivering irrelevant sessions still fails, just with better records. Our guide to building a learning culture covers the demand side, and blended learning design covers the mix most of these systems end up administering.
Selecting software first means automating whatever informal process already exists. Map request, approval, scheduling, attendance, records and reconciliation first — that map is your requirement document, and it costs nothing to produce.
Capture requiring every participant to log in on venue wi-fi loses to a paper sheet every time. Design it for the trainer standing in a room with twenty people waiting, and the record will exist.
Eligibility is decided at approval, before anything is booked, and cannot be applied retrospectively in either market. Making it a field on the request separates recovering cost from discovering the loss a quarter later.
They do not appear in the HR feed but do attend safety and compliance training on your sites. If their enrolment is a workaround, records will be incomplete precisely where completeness matters most.
Most employees only ever see the notification and the request form. Localising menus while leaving those in English gives you a system that looks regional in a demo and reads as foreign on the floor.
If chat requests still get actioned, chat remains the fastest route and people keep using it. Set a date, communicate it through managers, and hold it once the system has proven itself.
What you are building is a workflow with software attached, not the reverse. Organisations that get this right in Malaysia and Singapore do three unglamorous things: make the request short enough to complete on a phone, make attendance capture faster than paper, and decide funding eligibility before anyone books a date.
Everything else — dashboards, catalogues, integrations — is genuinely useful and genuinely secondary. Get those three right and records stay complete, claims stay recoverable, and employees use the system because it is the path of least resistance rather than because they were told to.
If compliance training is the bulk of what your system administers, our compliance training overview covers the record-keeping side in more depth, and our corporate training hub covers programme design once the administration is running.
Bring one real scenario — a claimable classroom session with an external trainer, mixed permanent and contract attendees, across both markets. We will run it end to end on the call, from request to claim pack.
Zainab is an experienced LearnTech leader with a strong track record of building and scaling digital learning solutions across the Middle East, Africa, APAC, the UK, and the USA. With deep expertise in Generative AI, capability development, and data-driven learning strategies, she has helped organizations modernize their learning ecosystems, enhance employee readiness, and deliver impactful, scalable L&D outcomes. Her work blends innovation with strategic clarity, enabling enterprises to adopt future-ready learning models that drive sustainable growth.
See how enterprises close skill gaps with AI. We'll email your brochure instantly.
Please enter your name, a valid email, and your phone number.
We respect your privacy. No spam — unsubscribe anytime.
Your platform overview is on its way. You can also download it right now.
Download Brochure











.png)
.png)
.png)
%20(1).png)
.png)







.webp)











.png)
.png)
.png)
%20(1).png)
.png)















Skillsoft is a global leader in corporate learning, providing digital training and education solutions to help businesses improve workforce productivity, reduce risk, and increase innovation.

FinShiksha provides a practical and industry-relevant approach to finance education, with courses designed by industry experts and delivered through interactive and engaging methods.

Wall Street Prep offers best-in-class financial training for aspiring finance professionals and corporate clients.

Udemy Business offers an unparalleled learning experience for organizations looking to upskill their workforce with over 155,000 courses taught by expert instructors.







.webp)








