Part 1 — ‘Legal engineering’: A Fragmented Landscape

The term ‘legal engineering’ is attracting renewed attention. Lawyers are building applications, law firms are recruiting technical specialists, legal technology companies are developing new roles, and emerging communities are articulating their own visions of the field. Generative AI has brought additional momentum. Activities that previously demanded specialist programming knowledge are becoming accessible to a wider group of people. The ability to demonstrate a working legal application has become an increasingly visible expression of technical capability.

Yet behind this growing visibility lies a fragmented landscape of roles and activities. The term ‘legal engineering’ is used differently by employers, people describing their work, technology companies and emerging professional communities. The activities described under it vary, as do the backgrounds of those performing them. Some accounts emphasise legal expertise and software development; others focus on technology implementation, workflow design or the development of systems used to deliver legal services. This fragmentation is not simply a matter of inconsistent terminology. It raises more fundamental questions about the nature of the field, the relationship between its different forms of practice and the extent to which any shared professional identity has emerged.

For employers, these questions have a practical dimension. What work needs to be done, which competencies does it require, and how can those competencies be identified, assessed and developed? Examining the work associated with the term ‘legal engineering’ can help clarify these questions without assuming that a single occupational title supplies the answer. The term ‘legal engineering’ may be gaining visibility faster than agreement is developing about what it describes. That tension provides the starting point for this series.

I. A field without a single centre

This series uses ‘field’ broadly to describe the activities, people, organisations and technologies associated with the term ‘legal engineering’. It does not assume that these constitute a single occupation or a profession. We examine how the work is distributed, whether recognisable occupational patterns are emerging, and how claims to professional status are being made. A unified occupation or profession is one possible outcome, rather than an assumed destination.

A. Similar titles and different work

The difficulty begins with the language used to describe the work. Consider the contemporary employment market. The research assembled for this project includes vacancies advertised under the title ‘legal engineer’, alongside roles described as legal technologist, legal solutions engineer and legal operations engineer. The employers include law firms, legal technology vendors and in-house legal teams.

The differences are not confined to titles. One employer may place considerable emphasis on understanding legal workflows and translating requirements into technology solutions. Another may describe a position centred on implementing a particular product. Elsewhere, the work may involve software development, systems integration, automation or the evaluation of AI-enabled applications. These descriptions cannot simply be treated as interchangeable. They arise in different organisational settings, with different responsibilities and expectations. A position within a legal technology vendor may serve different purposes from a technical role embedded in a law firm's practice or an in-house legal department.

Nor is it necessarily correct to assume that every role involving legal technology should be classified under the term ‘legal engineering’. The relationship between the occupational title, the activities performed and the professional identity of the person doing the work remains an important question. Nevertheless, the existing evidence establishes a significant point: work associated with the term ‘legal engineering’ is being described through multiple occupational vocabularies.

The exploratory evidence comprises an initial collection of 50 recruitment and narrative records: 44 recruitment records and six first-person or organisational accounts. It brings together the original 25-record dossier and a selected continental European component of 25 advertisements from 20 employers across seven countries, including Switzerland. The accompanying sources and method note preserves the record identifiers, source links, collection dates, duplicate controls and limitations. These records illustrate variation; they do not permit estimates of the prevalence of particular activities or qualifications across the employment market.¹ This distinction matters. The objective is not to manufacture occupational unity by grouping together every technologically oriented legal role. It is to understand the descriptions that employers and people doing the work actually use, including the differences between them.

B. Competencies and employer assessment

For recruitment, the implication is that a title alone provides an incomplete basis for judging suitability. The comparison needs to connect the activities and responsibilities of a role with the evidence of competence sought from candidates. Different educational and professional backgrounds may provide relevant expertise; their relevance depends on the work required. Job descriptions can reveal advertised expectations, but further evidence is needed to understand how employers assess candidates and how those expectations operate in practice.

In this series, capability describes what a person or team can do. Competencies are the knowledge, skills and behaviours relevant to performing particular work; competence means demonstrated ability to perform that work to an appropriate standard. These are working distinctions for the discussion, rather than universal definitions: different frameworks use the terms differently. Capability also remains useful when discussing what technologies enable people to do.

Three developments need to be distinguished: specialist appointments; competencies acquired within an existing role, including legal practice; and expertise distributed across a team. Growth in the ability to use or build technology does not, by itself, establish growth in a separate occupation. These distinctions guide the evidence examined throughout the series.

C. Occupation and profession

In this series, an occupation means a recognisable category of work involving related activities and capabilities. Software development, for example, can be identified as work without everyone doing it holding the same qualification or belonging to one professional body. A profession involves specialised expertise together with institutional arrangements for establishing competence, setting standards, accountability and recognition. Solicitors in England and Wales provide a familiar example: qualification requirements, professional standards and disciplinary arrangements support their recognised status.

The distinction is not simply between unregulated and regulated work: statutory regulation is not essential to every account of a profession. Nor does identifying an occupation mean that it will subsequently become a profession. Professionalisation is one possible development; related roles may also continue under different titles and professional affiliations.²

The first question is what work is being done under the title ‘legal engineer’ and related titles. Does it constitute a recognisable occupation, or several overlapping occupations? If an occupation can be identified, could it develop into a profession, under what conditions, and would that development be desirable?

The fragmentation also extends beyond recruitment. Law firms, technology companies, universities, training providers, professional communities, professional bodies and regulators approach the subject from different starting points. Some emphasise legal expertise; others foreground technological capability, process improvement, product development or practical software construction.

The continental European supplement broadens the comparison across seven countries, including Switzerland. Its advertisements describe workflow automation, AI adoption, customer implementation and software development. Part 4 examines the contrasts and the limits of the two collections in detail. These are contemporary occupational descriptions, distinct from historical legal-informatics evidence.³ Each perspective may illuminate part of the field. None should automatically be treated as an authoritative account of the whole.

II. Generative AI and the acceleration of visibility

A. What AI changes

The recent acceleration of interest in activities associated with the term ‘legal engineering’ cannot be separated from developments in generative AI. AI-assisted programming tools have made it easier for people to experiment with software development, automate tasks and produce working prototypes and applications. Activities once associated primarily with specialist developers can increasingly be attempted by people whose principal training lies elsewhere.

For lawyers, AI-assisted development changes some practical constraints. Historically, a lawyer seeking to translate professional knowledge into a working technological system might have needed to commission software development, collaborate with technical specialists or work within the constraints of existing products. A lawyer can now describe intended functionality, generate code, test an implementation and iterate on the result with forms of assistance that were not previously available.

Many of the problems addressed through AI-assisted building are familiar concerns of legal operations: understanding workflows, redesigning processes, reducing repetitive tasks and improving how legal work is organised and delivered. Lawyers building applications may therefore be addressing problems that people working in legal operations have approached through process improvement, technology adoption and collaboration with technical specialists.⁴ AI-assisted development can change how solutions are produced and who can build them. That change deserves examination alongside the continuity in the underlying activities. Understanding the work associated with the term ‘legal engineering’ requires attention both to these changing capabilities and to the contributions of those already analysing, organising and improving legal work.

B. Visibility and professional identity

The consequences extend beyond productivity. Building technology has become a visible means of demonstrating professional capability. Lawyers can publish applications, share demonstrations and present themselves as combining legal knowledge with technical execution. This visibility creates opportunities for new professional identities to emerge.

LegalQuants is one contemporary example. Its published account of the term ‘Legal Quant’ centres on lawyers who build technology to practise law. Jamie Tso’s founding essay describes configuring, evaluating and integrating AI into workflows, as well as building tools and collaborating with engineers. It connects these capabilities with professional differentiation and new business models.⁵ That account deserves serious examination. It describes a particular approach to practice and gives its participants a recognisable identity. But the emergence of such an identity raises a broader question: how far can an account centred on lawyers describe work distributed across different professional backgrounds?

LegalQuants provides one explicit contemporary account of a lawyer-centred technical identity. Its published materials establish how the organisation describes that identity; they do not establish how representative it is of the wider field.

Equally, it would be mistaken to dismiss AI-assisted building as a superficial activity. Some people building tools may use these tools primarily to generate prototypes; others may undertake considerably more extensive work involving system architecture, integrations, evaluation, deployment and maintenance. The distinction cannot be established merely by asking whether someone uses AI to write code. It requires examination of the actual work, its technical demands, the responsibilities assumed and the evidence of functioning systems.

C. A longer history

The historical point is equally important. Legal applications of symbolic logic, information retrieval, computational reasoning and expert systems long predate generative AI. These developments should not be retrospectively described as a single occupation bearing the label ‘legal engineering’. Nevertheless, they establish that the interaction between legal knowledge and technical systems has a substantial history.⁶

The historical record also extends beyond the Anglo-American examples often foregrounded in current discussion. Stockholm University traces organised law-and-informatics activity to 1968; the University of Namur records the establishment of its interdisciplinary Centre de Recherches Informatique et Droit in 1979; and Leibniz University Hannover dates its Institut für Rechtsinformatik to 1983, following an earlier research unit. These are histories of legal informatics and interdisciplinary institutions, not evidence that their participants belonged to a single occupation bearing the label ‘legal engineering’. They nonetheless widen the historical and institutional context in which contemporary claims should be considered.⁷

Generative AI has therefore not created the entire subject. It has altered the circumstances in which certain forms of practice become accessible, visible and professionally significant. Understanding that distinction is essential to understanding the present moment.

III. Plurality of practice and narrative dominance

A. Practice and prominence

Fragmentation creates a particular problem when a field begins to attract attention. Different communities may continue to perform different kinds of work, while public discussion may concentrate on a smaller number of influential interpretations. The distinction is between plurality of practice and narrative dominance.

Plurality concerns what people actually do, how they describe their activities and the professional settings in which they work. Narrative dominance concerns which descriptions acquire visibility, influence and apparent authority. The two are not necessarily aligned. A technically sophisticated community may be highly visible without representing the entire occupational landscape. A particular professional identity may be compelling and commercially successful without providing an exhaustive definition of the field.

This is not an argument against the communities advancing those identities. Emerging fields are often shaped by people who articulate new approaches, establish organisations and attract others to their work. Indeed, the ability to describe a distinctive practice coherently may be an important contribution. The analytical difficulty arises when the distinction between a particular model and the wider field becomes blurred.

B. Boundaries and entry routes

LegalQuants illustrates the question. Its account gives a name and a shared identity to lawyers with particular technical capabilities. The organisation explicitly describes an ambition to establish a new profession through training, community and assessment. This is an identifiable professionalisation project; its existence does not itself establish a recognised profession. Its community guidelines expressly distinguish admission from professional qualification or certification.⁸ The boundary it draws matters. Defining the ‘Legal Quant’ identity around lawyers privileges lawyer status within that account, although people from other backgrounds may contribute relevant expertise. Legal knowledge, software development, implementation and maintenance can be distributed across different people and organisations. Recognition centred on the lawyer who builds risks obscuring those contributions. Those people may or may not be covered by a defensible use of the term ‘legal engineering’. Their position cannot be resolved by adopting another community's terminology without examination.

Employer descriptions demonstrate why lawyer status should not be assumed. CMS’s Senior Legal Engineer AI / Development advertisement accepts degrees in law, computing, business or a related discipline, alongside relevant law-and-technology interest or experience. The project’s dated review of 25 continental European job advertisements records HVG Law as expressly stating that a specific legal or IT background is unnecessary. These examples establish different entry expectations, not their prevalence across the market.⁹

Even the route to qualification as a solicitor in England and Wales does not require an undergraduate law degree: the SQE route permits a degree in any subject, or an equivalent qualification or experience, alongside its other requirements. Educational background, legal competence and professional admission are distinct questions.¹⁰ The same caution applies to employers. A job advertisement establishes how a particular employer describes a particular position. It does not establish a profession-wide definition.

C. Definitions and authority

Contemporary narratives need to be examined with attention to the distinction between authority and representativeness. A credible examination of the term ‘legal engineering’ and the work associated with it must therefore consider not only what different sources say, but also the nature of the authority they claim.

An employer describes a vacancy. An individual describes their work. A professional community may propose an identity. An educational institution may describe a course of study. A standards organisation may publish a framework developed through a specified process. These are different kinds of evidence. Their claims should not be treated as equivalent merely because they use similar terminology.

These descriptions can also help organise the work they describe. A job advertisement turns an employer’s expectations into criteria for selection; an adopted assessment can make particular competencies count towards recognition. Alongside software that changes who can build, such documents and procedures can be examined through Actor–Network Theory as non-human actants (elements that help shape action): their significance lies in the relationships through which they shape participation and the distribution of work.¹¹

IV. Building technology and engineering systems

A working prototype can demonstrate initiative and technical understanding. It does not, by itself, establish how a system was specified, tested, secured, deployed or maintained. These wider responsibilities become important when an experiment enters sustained organisational use. Software engineering provides an independent comparator. The Guide to the Software Engineering Body of Knowledge (SWEBOK) describes requirements, design, construction, testing, maintenance, quality and professional practice as parts of a wider discipline. The comparison helps identify questions about dependability and responsibility; it does not impose a professional boundary on the work associated with the term ‘legal engineering’.¹² Part 6 examines these questions through the distribution of technical work and the evidence through which competence can be assessed.

V. Fragmentation does not necessarily mean failure

It would be easy to present the diversity of roles associated with the term ‘legal engineering’ as evidence of a profession that has failed to organise itself. That conclusion would be premature. Established disciplines accommodate different specialisations, professional backgrounds and forms of practice. Their participants do not necessarily share identical responsibilities or technical expertise. Variation can reflect legitimate differences in the problems being addressed.

The difficulty is determining whether diversity exists within a sufficiently coherent field, or whether apparently related activities have been grouped together primarily because of overlapping terminology. The answer has implications beyond semantics. For employers, occupational language may affect recruitment, assessment and career development. For people doing the work, it may influence how they describe their expertise and understand potential progression. For educators, it raises questions about the relationship between courses, qualifications and professional practice.

Yet the existence of these questions does not establish that a single professional standard is either possible or desirable. A common foundation, if one can be identified, need not require uniformity. Equally, similarities between different activities do not automatically justify bringing them under one professional designation.

The research must therefore resist two temptations. The first is to treat fragmentation as a problem that necessarily requires correction. The second is to allow fragmentation to become an excuse for avoiding difficult questions about professional knowledge, capability and recognition. A serious investigation needs to examine both the differences between forms of practice and the possible relationships connecting them.

VI. The question of professional recognition

Employers, universities, communities and training providers can help organise learning and recognition. Professional bodies and regulators are also key stakeholders: their existing competencies, standards and professional requirements need to inform consideration of assessment and recognition. Their contributions need to be examined in relation to the particular work and people they serve. A course, a professional identity and employer recognition are distinct achievements.

The practical issue is how people can gain confidence about relevant expertise while the work continues to vary. Part 7 examines the search for greater certainty in recruitment, development and the division of responsibilities. The final part considers possible arrangements for providing it.

VII. An inflection point, but towards what?

Current discussion of the term ‘legal engineering’ brings several developments together. It has a history extending beyond contemporary AI. Present-day activities associated with the term are described through multiple job titles and professional identities. Generative AI has accelerated the accessibility and visibility of some forms of technical practice. New communities are articulating increasingly confident accounts of what particular people do and why their capabilities matter. Yet the relationship between these developments remains insufficiently understood.

It is possible for a field to become more visible without becoming more coherent. It is equally possible for the emergence of competing identities to stimulate a more substantive discussion about professional practice. The direction is not predetermined. What matters is that the present discussion should not be confined to whichever account of what the term ‘legal engineering’ means happens to be most prominent. Lawyers who build applications have an important perspective. So do software engineers developing legal systems, technical specialists implementing legal workflows, employers recruiting for hybrid roles, and people whose work sits across established professional boundaries. Their activities should be examined on their own terms before conclusions are drawn about what connects them.

This first paper has deliberately stopped short of proposing a definition of the term ‘legal engineering’. That question requires more than an appealing formulation or a collection of examples. It requires attention to history, occupational evidence, technical disciplines and the different claims made by contemporary participants.

The enquiry also has a practical purpose for employers: to clarify the relationship between the work they need, the competencies they seek and the evidence through which competence is recognised. Alongside the series, the Institute’s consultation will explore these questions across recruitment, assessment, development and progression, including whether shared arrangements would be useful. The Institute will seek formal input from professional bodies and regulators on the relevant competencies and professional requirements. Training, community participation and employer recognition should be examined in relation to one another rather than assumed to coincide.

The subsequent papers will examine those subjects individually. The next instalment will return to the historical record, tracing relevant developments in legal logic, information retrieval, computational reasoning and legal technology before examining the later emergence of the term ‘legal engineering’.

The central observation of this first paper is that the work associated with the term ‘legal engineering’ is broader than its most visible contemporary narratives. Understanding its future requires understanding the plurality of practice that already exists.

Publication note

This series forms part of the Institute for Legal Engineering™ research programme. The Institute intends to develop professional standards, independent certification and training. The research examines the scope and justification for those activities; it does not assume that a single occupation, professional definition or certification model has already been established.

References

Legal Engineering, Occupational Evidence Audit and European Supplement V2 (project research record, 19 October 2026), original-record audit and European register. The register records access routes, duplicates, qualifications and limitations.Read the supporting sources and method →

CMS Hasche Sigle, ‘Legal Engineer (m/w/d) Automation’ (Germany; employer recruitment text, accessed 15 October 2026). Source

CMS Hasche Sigle, ‘(Senior) Legal Engineer (m/w/d) AI Adoption’ (Germany; employer recruitment text, accessed 15 October 2026). Source

CMS Hasche Sigle, ‘Senior Legal Engineer (m/w/d) AI / Development’ (Germany; employer recruitment text, accessed 17 October 2026). Source

Schellenberg Wittmer, ‘Legal Engineer 60–80%’ (Switzerland; employer recruitment text, accessed 15 October 2026). Source

LegalQuants, “About LegalQuants”, introductory statement and sections 1 and 3 (organisational self-description, accessed 17 October 2026). Source

Jamie Tso, “The Jane Street of Law: The rise of the Legal Quant” (15 January 2026), especially “The Jane Street Difference”, “Incentives and recognition” and “Clients: relationship business vs system business” (practitioner essay). Source

Layman E Allen, “Symbolic Logic: A Razor-Edged Tool for Drafting and Interpreting Legal Documents” (1957) 66 Yale Law Journal 833–879, especially 833 and the section “Application to Concrete Legal Problems”. Source

Bruce G Buchanan and Thomas E Headrick, “Some Speculation About Artificial Intelligence and Legal Reasoning” (1970) 23 Stanford Law Review 40–62, especially 40–41. Source

James A Sprowl, “Automating the Legal Reasoning Process: A Computer That Uses Regulations and Statutes to Draft Legal Documents” (1979) 4 American Bar Foundation Research Journal 1–81, abstract and note 20. Source

Stockholm University, Department of Law, “Centres and other units”, Swedish Law and Informatics Research Institute section (institutional retrospective, accessed 15 October 2026). Source

University of Namur, “1978–1992: essor et développement de la recherche”, CRID section (institutional retrospective, accessed 15 October 2026). Source

Leibniz University Hannover, Institut für Rechtsinformatik, “IRI – damals und heute”, especially “Entstehung” (institutional retrospective, accessed 15 October 2026). Source

IEEE Computer Society, Guide to the Software Engineering Body of Knowledge (SWEBOK Guide), Version 4.0a, publisher’s overview, resources and “SWEBOK Topics” (accessed 15 October 2026). The publisher identifies 4.0a as an update to Version 4.0. Source

Andrew Abbott, The System of Professions: An Essay on the Division of Expert Labor (University of Chicago Press 1988), Introduction, 1–3. Source

HVG Law, “Legal Engineer”, Rotterdam, employer-authored LinkedIn vacancy 4407734962, as recorded in Occupational Evidence Audit and European Supplement V2, E08 (checked 15 October 2026; the original link now redirects). Source

Solicitors Regulation Authority, “Degree and equivalent qualifications explained” (updated 23 July 2025) (accessed 17 October 2026). Source

Solicitors Regulation Authority, “What the SQE involves” (accessed 17 October 2026). Source

John Law, “Notes on the Theory of the Actor-Network: Ordering, Strategy, and Heterogeneity” (1992) 5 Systems Practice 379–393; author’s online reprint (2003), 2–3. Source

Solicitors Regulation Authority, ‘SRA Standards and Regulations’, including authorisation, conduct and disciplinary arrangements (accessed 17 October 2026). Source

Corporate Legal Operations Consortium (CLOC), ‘Practice Operations’, Core 12 framework (accessed 17 October 2026). Source

Legal Engineering, Occupational Evidence Sources and Method V2 (25 October 2026), collection composition, linked source index, duplicate controls and source-access limitations.Read the supporting sources and method →

LegalQuants, “Community Guidelines”, membership and assessment sections (accessed 25 October 2026). Source