Pure as a research truth engine — showcasing in the age of AI

From display board to truth engine

Research showcasing has changed. The question institutions face is no longer whether their research is visible to the right human readers, but whether it is discoverable to the algorithms and AI systems that increasingly stand between them and the audiences they need to reach. This shift has profound implications for the repositories, department pages, faculty directories and Research Information Management Systems (RIMS) that underpin their institutional showcasing efforts.

Classic showcasing – as exemplified by a mature RIMS – is the practice of presenting research expertise, activity and organizational strengths through a governed, public-facing portal. The aim is that academic collaborators, industrial partners, funders, journalists and members of the wider public will use the site to find a reliable, cohesive and up-to-date account of an institution's work. Elsevier’s Pure has helped set the standard here, signaling a clear advance on the early 21^st century experience of trawling through a scattered mix of departmental pages and out-of-date biographies in search of the right research information. The Pure Portal gives institutions a single, branded public window built on validated data rather than manually maintained pages. Similarly, Pure’s Community module extends this idea to consortia and innovation alliances, allowing a group of institutions to present a coordinated picture of their collective capabilities. While these classic showcasing features retain their value, the last five years have seen a significant change in the way they are perceived. The question is no longer about how you showcase, but who and what you showcase to.

Classic showcasing was developed for a world where the primary audience consisted of individuals browsing the internet via traditional search engines. Today, a rapidly growing share of an institution’s "audience" is comprised of AI systems parsing structured data to answer questions from those same users. In other words, the search experience is now proactively mediated by tools that act as a kind of digital concierge, effectively moving the burden of complexity from the user to the potential endpoints of their query. In this world, it is no longer viable for a RIMS to function simply as a display board, however sophisticated it might be. Instead, it must function as a research truth engine: an authoritative source that AI search can find, establish as credible and then cite.

The data foundation: control of your data is control of your story

The core use case of any RIMS has always been to serve as the single, authoritative source of truth for an institution's research activities and outputs. While the advantages of RIMS ownership are well known – reduced administrative burden, cleaner compliance, more robust impact reporting, more reliable financial auditing, better institutional risk management, a stronger evidence base for strategic decision-making – this cohesive data foundation now plays a key role in determining whether or not an institution is discoverable. Unlike the pre-AI world of traditional keyword-based search engines, where a RIMS data foundation was primarily an internal-facing asset, AI answer engines increasingly require a single authoritative destination like a public RIMS portal. This provides a stable web address an engine can point to and cite when grounding an answer in fact, the digital equivalent of a primary source. If an institution’s research presence is spread across multiple disconnected systems, then it risks something that is potentially even worse than obscurity – losing control of its own research narrative. If a research organization is unable to tell a joined-up story, third-party web scrapers and aggregators will fill the information gap with whatever fragments they can piece together, following their own agendas rather than the institution’s.

The depth of Pure's data model can ensure institutional narratives are both cohesive and compelling. As a RIMS, Pure’s advantage is less about how many sources it pulls in than how rigorously it reconciles data that might be naturally fragmented across an organization. Faculty records from HR, grant data from finance, patent activity from the technology transfer office and publication records from the library are all drawn into a single, tightly reconciled schema rather than sitting as loosely linked references. Where the same output shows up in multiple sources – say, a publication indexed separately in Scopus, Crossref and an internal repository – Pure's deduplication tooling automatically merges these into one verified master record. Disambiguation – the process of correctly identifying and separating individuals or organizational entities with the same or similar names – is a problem that automated systems either get right through structured identifiers or get wrong at scale. A funder database or an AI tool that is unable to tell which "J. Smith" co-authored an article will not credit the right person, or the right institution, for this work. Pure's disambiguation capabilities – derived from a structural relationship with Scopus, not an API feed – help determine whether an institution's research data actually gets attributed correctly once it leaves the building.

This same governed data foundation also supports the drive to find suitable collaborators. For example, a potential corporate partner, effectively doing due diligence on an institution's expertise before committing any money, will find a validated, disambiguated record a far more credible basis for that conversation than a set of unaligned departmental web pages, or even the public-facing site of a RIMS that lacks the right underlying data sources, algorithmic sophistication, or persistent identifiers. As industry funding becomes a proportionately larger share of the money in play in the research ecosystem, this credibility starts to matter as much as the compliance and reporting benefits that the RIMS data foundation was originally built to support.

Discovery: the pull strategy

A strong data foundation supports a coherent public portal that helps an institution get found, but relying exclusively on this approach is a dangerously passive strategy. In today’s hyper-competitive academic environment, the traditional model – publish the page and wait for a bot, journalist, academic or corporate partner to come across it – is no longer sufficient. Pure's architecture is built to make that wait shorter and more productive. A good example of this is the Pure Portal itself, structured in a way that suits how large language models actually parse information. Automatic schema markup and entity-rich concept indexing mean the Portal is eminently findable. Similarly, open-access full-text hosting on the Portal enables AI engines to crawl complete papers rather than abstracts alone, providing a structural advantage over custom-built or Content Management System-based sites that were not designed with machine readers in mind.

Embedded in Pure, the Fingerprint Engine addresses a different, related problem: being found by the right query. While being found by the wrong query might not seem like an issue, as long as there are an acceptable number of hits among the misses, irrelevant traffic can actively undermine a university's reputation by signaling poor data quality, creating accidental misrepresentation and generally frustrating potentially valuable partners. The main difficulty is that R&D professionals, Strategic Partnership Managers and Technology Scouts increasingly search by problem, not by job title or department. Pure's Fingerprint Engine helps because it uses natural language processing to index weighted semantic concepts across a researcher's entire body of work. A company looking for materials-science expertise can be matched on the underlying problem itself, rather than being required to know in advance which department or individual they are looking for. As industry becomes a larger source of research funding relative to government grants in many areas of the world, this distinction matters more than it used to.

This same trend is shifting the overall focus of research toward ROI-oriented applied problems and away from more theoretical “blue sky” work. This is bad news for universities that are heavily invested in basic research and need to derive academic, reputational and commercial value from it. While applied research can be presented as problem-led discovery – a company searching for materials science expertise will often be looking for a match to a well-understood and immediate challenge – this is not usually the case for curiosity-driven work. The Fingerprint Engine can help here by ensuring researcher profiles are built on a foundation of core concepts (e.g., Quantum Entanglement, Stochastic Processes), mapping their fundamental capabilities. In doing so, it can support both ends of the research spectrum by highlighting the conceptual foundations that connect diverse areas of inquiry.

A further discovery gap exists at the research collaboration level within consortia, regional innovation hubs and public-private co-labs. Research alliances are a rapidly growing trend in global higher education, with universities seeking to grow collaboration, optimize cross-border funding mandates and share the rising cost of large-scale science. The problem is that while an institution can have clean, well-governed data for every one of its own researchers, it can still fall short because its records were never built with cross-organization discovery in mind. There is often no shared schema for a partner to plug into, so the alliance ends up presenting itself a loose federation rather than a single capability – potentially a less compelling proposition for funders and potential partners. Pure's Community module is built specifically to support this use case. It lets research alliances, innovation clusters and national programmes present a combined portal with a shared, disambiguated structure underneath it, without any one institution losing control of its own data. Most single-institution RIMS deployments are simply not designed to handle this challenge, being built to support discoverability at the researcher level, but not the coalition one.

Proactive communication: the push strategy

While discoverability is crucial, an advanced Research Information Management System like Pure can also be used as a launchpad for targeted outreach – not so much a static directory as an active intelligence and matchmaking tool. Its synchronization processes run in both directions: Pure draws in validated publication and citation data from Scopus to keep the institutional record current, while also exporting content outward – publications, affiliations, even non-traditional outputs like preprints and software – directly into researchers' ORCID records. This matters because major global ranking bodies like Times Higher Education and QS, as well as international funding agencies, already treat Scopus data as a baseline source of truth.

The Fingerprint Engine's semantic indexing also has a role to play in supporting an outbound showcasing strategy. It can be used to proactively match the semantic profiles of incoming funding opportunities or collaboration requests against the indexed expertise profiles of the faculty members most likely to be a good fit. Because the Engine relies on controlled thesauri and deep semantic extraction rather than basic keyword matching, it can identify genuine conceptual fits even if the incoming request uses slightly different terminology than the faculty member's past publications. This streamlines much of what is often a manual, relationship-driven business development process through automated matching.

Pure also supports richer impact storytelling by providing both quantitative evidence (e.g., patent citations, policy citations, SDG alignment) and qualitative evidence (e.g., news articles, quotations, letters of support) – proving how much institutional research has been noticed and explaining how and why it has changed the real world. As public funders become increasingly keen to demonstrate how the work they support has influenced society, the economy, health systems or national culture, this combination is rapidly becoming the narrative gold standard.

Pure can enhance institutional-level storytelling by providing institutional communications teams with a custom evidence feed to work from. The RIMS integrates with PlumX Metrics, Elsevier's own altmetrics suite and, via an API connector, Altmetric's Explorer. Together, these tools surface citation counts alongside broader attention signals such as news coverage and social and mainstream media mentions – all tied back to a validated research record. This gives communications teams a headstart in identifying which outputs are gaining traction and deserve a more proactive push.

The tension between openness and security

Whether pushing or pulling, the practice of showcasing does not take place in a vacuum. Institutions are being asked to be more open, more structured and more machine-readable at exactly the moment when research itself has become more geopolitically and commercially sensitive. Getting the balance wrong in either direction comes at a cost. If an institution is too closed, then it risks invisibility to the AI-based systems that, as we have seen, are now mediating discovery. However, if an institution is too open, then there is a danger of exposing research relationships, research pipelines and even personal data to competitors or bad actors.

This is a societal problem, but it also has quite specific implications for RIMS and the way an institution projects its achievements into the outside world. For Pure, governance and access control have always been primary design concerns. The system enables institutions to apply role-based access so that the public-facing Pure Portal shows a filtered, appropriate view of the underlying data, while more sensitive relationships – an early-stage industry conversation, a not-yet-public grant or a project still working through internal review – can be kept out of the public-facing layer until an institution is ready to disclose them. Because of these capabilities, Pure can also be integrated into institutional enterprise AI systems – often prone to insularity because of their commitment to “walled garden” integrity – to act as a governance layer between the sovereign AI environment and the outside world.

This dexterous handling of exposure and disclosure is also important to the successful management of compliance. In Europe, where compliance regulations historically catalysed the development of the RIMS, institutions intent on publicly archiving their academic achievements must also ensure they respect the "right to be forgotten" defined by the General Data Protection Regulation (GDPR). This pressure is mirrored in Asia, where frameworks like Japan's Act on the Protection of Personal Information (APPI) impose strict consent structures over how a researcher’s personal and historical data can be transferred or publicly displayed. In the US, which historically approached compliance in a highly decentralized way, measures like National Security Presidential Memorandum-33 (NSPM-33) now require federally funded researchers to disclose research support, appointments and affiliations so agencies can identify overlaps in funding or commitment. Similarly, the Family Educational Rights and Privacy Act (FERPA) adds a parallel layer governing how institutions handle student and researcher education records. The key takeaway here is that the same information cannot be treated one way for compliance purposes and another for public-facing visibility. This combination makes a researcher's profile an institutional asset that needs the institutional-grade governance that Pure can readily provide.

Pure – the data is the destination

AI tools have shifted the modus operandi of research showcasing from visibility to discoverability, but not everyone is keeping pace. By mediating search, these tools have pushed the burden of complexity back to the institutions themselves. This means that a university with data that is messy, buried in unindexable PDFs, or lacking machine-readable structure, will simply be bypassed. The “display boards” that leading RIMS provide, while based on validated data, are no longer enough.

While Pure provides a best-in-class public-facing website, it also functions as a research truth engine: an authoritative source that AI search can find, establish as credible, understand in context, and confidently present to the end user. While Pure has always served as the definitive record of an institution's research outputs, the data is now the destination. Institutions that do not control that data will not control their research narrative.

We live in an age where, driven by compressed digital feedback loops and predictive AI systems, perception and action are becoming the same thing. The research an institution produces may be strong enough to win a grant or an industry partnership, but unless it can be discovered, it will only deliver a fraction of its academic, commercial and reputational value. At a time when many universities are under severe financial pressure, obliged to do more with less, harnessing the full potential of institutional research outputs is vital. Pulling and pushing, open and secure, Pure has been optimized to help you realize his promise.