Chapter 2 · Guide

Context

The three types of organisation the taxonomy serves and the four areas of need it is designed for: delivery, knowledge, pipeline and people.

2.1Three types of organisation

The noslegal taxonomy is designed to support three types of organisation:

  • Legal services providers — lawyers and others who practise law and provide related services. This includes traditional professional organisations such as law firms and barristers' chambers, not-for-profit providers such as law centres, and alternative legal services providers (ALSPs).
  • In-house legal services — the legal departments of businesses and larger public bodies. These do some legal work directly and manage the sourcing of other work from legal service providers. They operate under particular pressure to demonstrate value, control spend and manage an ever-broadening range of legal risk, often with tight resources.
  • Legal technology companies provide software and, in some cases, information used either by the first two groups or directly by end users such as individuals and small businesses.

The taxonomy is not specifically designed for legal authorities (courts, regulators, law enforcement), legal research, teaching, or theoretical and academic law. These areas tend to make jurisdiction-specific and legal-conceptual demands that would undermine the practical, internationally neutral focus of noslegal. This is not a criticism, it's just a choice we have made in order to stay focused. If you're working in these areas, or anything else, you are of course free to use, modify and extend noslegal if you do find it relevant.

2.2Four areas of need

Legal organisations rely on structured information about their work across four broad areas. These areas overlap and interact — and the value of a shared taxonomy is precisely that data generated in one area becomes useful in the others.

  • Delivery. Scoping, planning, pricing, executing and managing the work, including adjustment of plans and pricing as matters evolve.
  • Knowledge. Capturing, structuring and making available insights and materials that support legal work, drawing on the organisation's collective experience as well as external resources.
  • Pipeline. Understanding, anticipating and influencing what work is upcoming or available. Within a legal services provider, it includes marketing, business development, credentials and client relationships. For in-house legal teams, it includes forward planning, resourcing, managing the flow of work from the business, and handling legal services provider relationships.
  • People. Understanding and communicating what experience and skills an organisation's people have so that work can be resourced effectively. Also determining what's missing so that gaps can be filled by recruitment, training, secondments or in other ways.

These four areas overlap and interact in ways that make it desirable for the higher-level concepts used in each to correspond.

Examples of how these needs benefit from shared concepts

A law firm will ideally target its marketing and sales activities (Pipeline) on areas which can generate significant amounts of worthwhile work (Delivery), make relevant learnings as to how to do so effectively and profitably (Knowledge) and generate credentials (People) which can support marketing and sales (Pipeline again) and help those doing the work in future (Delivery again and People).

An in-house team delivering legal work (Delivery) generates know-how and financial data (Knowledge) that can inform how similar work is planned, resourced (Pipeline), priced and managed in future (Delivery again), including how it is staffed and how career development is supported (People).

The core functions in these four areas of need are often handled by separate groups within many legal services providers, each with its own characteristic software applications. The result is that the same underlying work tends to be described differently across functions — making it difficult to connect insights, identify patterns, or reuse what has been learned.

Using a shared high-level taxonomy helps to address this. It avoids the constant mapping and remapping exercise that arises when different functions, or different organisations working together, maintain their classifications independently.

Needs will typically diverge at lower levels of detail. For example, a knowledge management function will typically have more need for more granular legal concepts and sub-concepts than a team focused on financial reporting or business development. But this can be handled by extending the taxonomy to meet such specialist needs without disturbing the shared higher-level concepts on which interoperability depends.

2.3Approaches to information which noslegal enhances

Legal work has always depended on finding and reusing information generated elsewhere. Digital technology has already extended this considerably in recent decades. New varieties of AI are now making new approaches feasible. All of this benefits from a reliable, consistent conceptual skeleton.

Well-established approaches, enhanced with a strong taxonomy, include:

  • Search: finding something you can describe. Taxonomy improves precision by ensuring that what was classified under a given concept can be found under that concept, not a near-synonym or a different organisational convention.
  • Alerts and dashboards: flagging developments and keeping views current based on defined rules. Taxonomy provides the stable categories on which those rules depend.
  • Rule-based assembly: producing reports and documents based on predefined rules and selections. Taxonomy supplies the selection criteria.

These are useful but can be limited by the active engagement they require and a tendency to return too much. Newer approaches using non-deterministic technology such as machine learning and language models include:

  • Contextual retrieval: returning not just what was asked for but what is likely to be helpful. Taxonomy enables pre-filtering of the corpus before a model processes it, improving both relevance and performance (see section 6.2).
  • Question answering: addressing queries directly rather than pointing to search results. Taxonomy provides some meaningful structured context to help a model identify authoritative sources and assess coverage gaps.
  • Generative assembly: producing reports and other outputs from loosely specified requests. Taxonomy anchors what the model draws on, reducing the risk of plausible but poorly grounded outputs.
  • Process-linked surfacing: presenting relevant information at the step the user has reached, or raising something they ought to address. Taxonomy, combined with process elements (section 4.7), is what links the step to the information.
  • Agentic AI: taking action without seeking confirmation. Taxonomy can be used to constrain the scope within which an agent operates. It can also provide a classification layer against which its actions can be audited.

Each approach has limits and risks, and in practice they work best in combination. A shared taxonomy is what makes that combination practical. Different approaches can operate over the same classified corpus without each needing its own organisational scheme.

2.4Examples of use cases

Across the four areas of need noslegal is designed for, the approaches just listed can be used in many ways. This table gives just a few examples with a deliberate mix of ones for legal services providers and legal departments.

At matter levelAcross a portfolio
Delivery What should the plan be for this matter — work, deliverables, assumptions, budget? Where are the high-risk areas? How might fixed pricing work out here, based on previous matters? At this stage, what are the things that need to be covered? What kinds of work have we handled for this client or business unit? What types generate the greatest cost or risk? Where do we see recurring overruns we could manage more effectively?
Knowledge How should I approach this issue? What has our organisation said or decided about this before? How has it worked out in practice? What context can we share for future work? How proportionate and up to date are our knowledge resources relative to the work we actually do? Where are the gaps?
Pipeline (To help with sourcing) Have we faced matters like this before — who did them and how did it go? (To help win work) What experience can we point to for this pitch review? What patterns are we seeing across our work? Where is demand growing or declining? Where are we under- or over-resourced? (Provider) What are our realisation rates across different kinds of work and where can we improve? (In-house) Where is demand from the business increasing and how should we plan for it?
People (To help with resourcing) Do any of our lawyers have capacity and experience in this kind of matter? (To help with people development) Are we spreading experience widely enough to support development? Where should we focus training and in-context help?