A previous article on The Cove, "What if the Work Is the Training?" (Ashford-Rowe, 2026), argued that the Army should treat authentic workplace activity as a legitimate source of learning and, where appropriate, credible evidence of capability. It proposed a shift in thinking:

The workplace can generate the evidence; the assessment system must assure the judgement.

That argument leaves a practical problem unresolved. Evidence generated continuously through work – for example, a brief, an after-action review (AAR), feedback from a subordinate, or a difficult decision under pressure – is by default scattered, ephemeral, and unassured. A conversation in a headquarters does not automatically become defensible evidence against a standard. Something has to capture it, curate it, connect it to a standard, and make it available for judgement and moderation.

This article argues that the ePortfolio is that mechanism: not a digital filing cabinet for certificates and course completions, but a structured, learner-owned, assessor-accessible record that turns the ordinary residue of authentic work into evidence an assessor can actually use, and the connective infrastructure through which formal learning, work, feedback, reflection, judgement, and recognition remain connected over time.

What is an ePortfolio?

For readers outside training and education, an ePortfolio is best thought of as a digital, organised, continuously updated collection of evidence of what a person has done, learned, and can do. The individual holds it, but it is structured so someone else can share, review, and judge it. The analogue equivalent is familiar in several trades and professions: an artist's folio; an architect's set of plans; or the folder a tradesperson keeps of certificates, photographs of finished work, and a logbook of hours. Those folders work, but they are static and usually assembled only retrospectively, pulled together when applying for a licence, long after the work itself and with no mechanism for checking evidence against a standard, or for a second reviewer to test whether a judgement was reasonable.

ePortfolio platforms have matured considerably over the past two decades, moving from little more than online CVs into structured systems that timestamp and attribute evidence, prompt reflection, map entries to named competencies, and support a visible layer of assessor comment and moderation. That maturity is why accrediting and professional-registration bodies in teaching, nursing, medicine, engineering, and other regulated professions increasingly accept and sometimes require ePortfolio evidence to demonstrate competence: a legitimate, auditable evidence base, not merely a training record.

From concept to mechanism

It is one thing to accept, in principle, that authentic work can generate credible evidence of competence. It is another to build a system that can reliably capture and present that evidence in a form an assessor can trust. Without a deliberate mechanism, three things tend to happen: evidence is lost, because nothing durable remains once an activity ends; evidence is unverifiable, because a learner's account of what they did is hard to distinguish from what they wish they had done; and evidence is disconnected from standards, leaving an assessor substantial interpretive work before it becomes usable.

An ePortfolio, properly designed, addresses each of these. It gives the learner a place to capture performance-relevant information (Ajjawi et al., 2026) close to where it occurs; it structures that capture so evidence can be attributed, dated, and contextualised; and it requires evidence to be explicitly linked to the standard it is claimed to demonstrate.

What an ePortfolio actually needs to do

A workable ePortfolio needs to do at least four things well, aligned with established portfolio design principles (Klenowski, 2002) and what makes workplace evidence genuinely authentic (Ashford-Rowe, Herrington, & Brown, 2014).

Capture evidence close to the work, not after it. The further capture drifts from the activity, the more it becomes reconstruction rather than record. Learners need low-friction ways to attach orders, briefs, AARs, supervisor comments, or short reflective notes near the point of performance.

Require reflection, not just artefacts. A saved document is not, by itself, evidence of understanding. Following Carless and Boud's (2018) concept of feedback literacy, a workable structure prompts the learner to interpret the artefact: what standard it demonstrates, and how they responded to it, developing the capability to notice and interpret, not just store evidence.

Map evidence explicitly to standards. Every entry should be taggable against the competency it is offered against, letting an assessor ask not “did something happen” but “does this support a judgement of competence?”

Make judgement, not just storage, but visible. An ePortfolio needs a visible assessor layer: professional judgement recorded against evidence, moderation comments alongside it, and a defensible audit trail. Judging portfolio evidence well is closer to what Boud and Dawson (2023) describe as teacher feedback literacy than to conventional marking. Without this, an ePortfolio is a repository; with it, it becomes the connective tissue between authentic work and assured competence.

Guarding the line between feedback and evidence

The earlier article warned against turning every workplace interaction into an assessment opportunity, drawing on Bearman's (2026) analysis of feedback culture: if every exposed mistake risks a formal judgement, learners will rationally stop exposing them. An ePortfolio can protect that distinction or erode it, depending on how it is designed. If every entry is automatically visible to an assessor as potential evidence, the tool becomes a surveillance instrument, and the behaviours it is meant to encourage – noticing, questioning, admitting uncertainty – will quietly disappear.

A workable model needs a genuine developmental space, visible only to the learner and perhaps a mentor, alongside a deliberately curated evidence space the learner chooses to submit for assessment. The learner, not the system, decides when performance-relevant information has matured into evidence. That is the distinction that lets an ePortfolio support learning and assessment without collapsing one into the other.

The ePortfolio as a moderation artefact

As argued previously in relation to assessment within the military environment (Ashford-Rowe, 2018), the Army does not need to standardise every experience through which capability is demonstrated, but it does need sufficient consistency in the basis on which competence is recognised: common standards, evidence requirements, assessor judgement, moderation, and quality assurance. An ePortfolio carries that weight because it makes evidence and judgement visible and comparable, even where the underlying experiences were not.

Two officers may submit entirely different evidence – for example, one from a planning exercise and another from a live operational task, but if both are structured against the same standard, an assessor panel can compare judgements, not experiences. That is what moderation actually requires: not identical inputs, but a shared, examinable basis for comparing outputs, a more repeatable calibration tool than periodically comparing notes on a residential activity.

Where this goes wrong

An ePortfolio is not a solution by itself, and the ways it can fail echo long-standing critiques of portfolio-based assessment (Driessen, 2017).

It can become a compliance exercise, rewarding volume over quality, reflecting the fixation on inputs and outputs seen elsewhere in intensive and non-traditional study (Lodge & Ashford-Rowe, 2024). It can be gamed: a polished retrospective narrative is not evidence generated at the time, so contemporaneous capture matters. It can reproduce workplace unevenness where an engaged supervisor helps build a stronger portfolio for reasons unrelated to competence, so assessors need licence to seek supplementary evidence rather than judge only what was submitted.

And it can become another administrative burden on existing reporting; if it does not visibly reduce residential days or duplication, adoption will be superficial, precisely the pattern in sectors with a mature ePortfolio history, where institutional and technological factors, more than pedagogy, have determined whether an initiative survives its pilot (Coffey & Ashford-Rowe, 2014).

From portfolio to learning ecosystem

The larger opportunity is to treat the ePortfolio not simply as an assessment instrument, but as the organising infrastructure for a learning and assessment ecosystem in which formal learning, authentic work, feedback, reflection, judgement, and recognition are deliberately connected with the individual and the capability standard at the centre, and with courses, workplace experience, and assessment as different ways of developing or demonstrating what is required.

That suggests a capability lifecycle rather than a course sequence: an individual begins with a defined standard and whatever credible evidence already exists, which identifies what remains to be learned; formal learning is used where it adds value; workplace activity develops and demonstrates capability; selected evidence is curated against the standard; an assessor judges; and moderation provides confidence in it. The portfolio continues with the individual rather than ending when a course does.

At system level, six elements interact: standards defining capability and acceptable evidence, formal learning, a workplace generating authentic evidence, a portfolio for capturing and curating it, assessor judgement, and assurance providing calibration across dispersed learners. The ePortfolio connects these layers; it does not replace them, and different learners may follow different pathways while held to the same exit standard. One boundary is important: the ePortfolio must never become the standard. Completing a portfolio is not competence, and accumulating entries is not evidence of capability by volume, the standard remains external to the portfolio, which organises evidence against it, while the assessor judges and the assessment system assures that judgement.

A particular opportunity for the Reserve

Even accounting for those risks, Army Reserve is a context where the upside is particularly concentrated, given the competing demands on Reservists' time. A Reservist moving between civilian employment, unit activities, and periodic training accumulates performance-relevant information in fragments, often months apart, in ways that would otherwise never come together into a coherent evidence base. An ePortfolio that a Reservist can update opportunistically – for example, after a planning activity, a course, or a civilian workplace responsibility with transferable relevance – turns that fragmented experience into something an assessor can review as a whole, rather than recreated from scratch at the next residential block.

What a workable model might look like

This architecture does not require a single enterprise system before anything can begin; designing the whole ecosystem at once would risk creating technology and process before the Army has learned what evidence, judgement, and assurance it actually requires. A staged implementation is more realistic:

1. Start with one capability or course where authentic evidence is plausible; an assessment-only pathway is a natural pilot.
2. Define what counts as evidence for that capability, mapped explicitly against the relevant standard.
3. Use existing evidence diagnostically to identify what has already been demonstrated and what remains necessary.
4. Design the portfolio around the capability standard and evidence requirements, not around completing the portfolio itself.
5. Separate the developmental space from the evidence space, and make that separation visible to learners.
6. Require contemporaneous capture wherever possible: dated, attributed, and witnessed where appropriate, rather than retrospective narrative.
7. Build a visible assessor and moderation layer into the same tool, not a separate process bolted on afterwards.
8. Train assessors specifically in judging portfolio evidence; this is a different skill from marking a scripted assessment activity.
9. Review after one cycle: is evidence quality improving, is assessor confidence increasing, and has anything actually been removed from the training burden as a result?

If the pilot works, scale should come through common standards, evidence rules, assessor development, and assurance rather than forcing every unit to reproduce identical experiences. Technology should make evidence portable and visible, but the scalable asset is ultimately the shared basis for professional judgement.

Scaling that technology, however, raises questions no pilot can defer indefinitely. Any enterprise ePortfolio will need interoperability with existing Defence systems rather than becoming a parallel record nobody trusts. It would also require a stable identity model so a member's evidence travels with them across postings, categories of service, and even between Services rather than lapsing with a unit-specific login; cyber security controls and classification handling appropriate to evidence that may reference locations, capabilities, or personnel; data sovereignty arrangements, particularly for a commercial or cloud-hosted platform, that keep the record under Australian and Defence control; and genuine portability, so evidence built over a career is not stranded in a system tied to one posting or Service and can support recognition of prior learning into civilian employment at transition. None of this is a reason to delay a small pilot, but a pilot that ignores it will produce a proof of concept that cannot scale.

A longer-term possibility

As standard-evidence mapping matures, a further possibility is worth naming honestly, even though it sits beyond an initial pilot. Well-designed AI could, over time, assist in determining training outcomes, including recognition of prior learning, by supporting at scale the evidence-to-standard mapping this model requires assessors to do by hand, combining machine capability with human expertise, transparency, and control – not replacing professional judgement (Buckingham Shum et al., 2023).

Recognition of Prior Learning is often constrained less by whether relevant evidence exists than by the labour of finding it, mapping it against every current standard, and presenting it in a form an assessor can act on. An ePortfolio holding structured, tagged evidence accumulated over years is exactly the kind of record AI could search, cross-reference, and pre-map against a capability framework, surfacing where evidence already satisfies a standard and where a gap remains. For a Reservist carrying years of fragmented evidence and limited time, that difference is between RPL being theoretically available and practically usable.

The same boundary applies with more force here. An AI system can triage, surface, and pre-map evidence; it cannot make the judgement. Treating it as though it can would begin to collapse the assurance the assessment system exists to provide. Any AI-assisted recommendation must remain a visible, challengeable input, and its reasoning must be transparent enough for moderation to examine as closely as a human judgement. This would strengthen the assurance chain rather than shortcut it.

The record, not just the argument

The first proposition was that Army need not always take people away from work to develop and assess capability. To operate at scale, Army needs infrastructure connecting learning, work, evidence, reflection, judgement, and recognition over time without turning every workday into a test.

The ePortfolio, properly designed and deliberately bounded, can provide that infrastructure, not because the technology is new or clever, but because it can make that connective tissue visible, portable, comparable, and assurable. It is not another training administration tool but a persistent capability record within a broader Army learning ecosystem, tracing back through work on what makes workplace evidence authentic rather than merely convenient (Ashford-Rowe, Herrington, & Brown, 2014), the sector's own uneven experience of building ePortfolio systems (Coffey & Ashford-Rowe, 2014), and the argument that education has too often measured learning by its inputs and outputs rather than the process that produces it (Lodge & Ashford-Rowe, 2024). What changes here is only the setting.

The portfolio does not demonstrate competence. The evidence does. The assessor makes the judgement. The assessment system assures that judgement. If the work is the training, the ePortfolio is how Army makes that defensible at scale.

Still Interested?

Why not also read “Upside-Down Man of Performance Appraisal Reporting” by Mal Brick.

Explore Cove+ short learning courses on related professional military education topics.