Modern law firms run a stack of software that falls into a handful of core categories: practice management, document drafting and contract review, e-discovery, research and citators, billing and payments, and a growing layer of AI tools that cut across all of them. Which categories a firm actually deploys depends on its practice mix. A litigation shop leans hard on e-discovery; a transactional firm lives in contract lifecycle management.
The categories matter because each one carries its own procurement questions, its own data-handling risks, and its own professional-responsibility exposure. Buying the stack piecemeal without understanding how the pieces connect is how firms end up with overlapping contracts, unmanaged data, and tools nobody uses. This explainer maps the categories in the order work actually flows through a firm.
The starting point is definitional. As Cornell Law School's Legal Information Institute explains in its Wex glossary, legal means anything having to do with the law, and in everyday use refers to something that conforms to it. Legal technology, then, is software built to support work that has to do with the law — and the stack below is organized around the distinct jobs that work requires.
What does practice management software actually do?
Practice management platforms are the operating system of the firm. They hold client and matter records, calendars, deadlines, conflicts information, and task assignment. In many firms they also anchor time capture, so every hour worked attaches to the right client and matter code.
The category matters for compliance as much as for convenience. Court deadlines, conflicts checks, and trust-accounting records are professional obligations, not preferences. A firm that runs deadlines in individual lawyers' heads has a risk problem that no other tool in the stack can fix. Practice management is usually the first purchase a growing firm makes, and often the one other tools must integrate with. We covered a connected angle in State AI Laws Took Effect January 1: New Documentation Duties for Legal and Compliance Teams.
Which tools handle documents and contracts?
Document work splits into two related but distinct categories. Document drafting and assembly tools let lawyers build pleadings, agreements, and letters from templates, so routine language stays consistent and precedent gets reused deliberately rather than by copy-paste habit.
Contract lifecycle management, or CLM, covers the full arc of a contract: intake, negotiation, approval, signature, storage, and renewal tracking. Transactional teams use CLM to see obligations and key dates across a portfolio rather than one agreement at a time. AI contract review tools sit alongside this category; vendors claim they can flag risky terms and deviations from standard positions, and firms evaluating those claims should read them as vendor statements, not verified findings. For a closer look at that evaluation problem, see AI Contract Review Tools: What Vendors Claim They Do and Contract Lifecycle Management Explained: What CLM Platforms Do Across the Contract Lifecycle.
How does e-discovery fit into the stack?
E-discovery is the litigation end of the stack. It covers the preservation, collection, processing, review, and production of electronically stored information — emails, chats, documents, and other digital records that must be exchanged in litigation. The workflow typically starts with a legal hold, which directs custodians to preserve relevant material, and ends with a production set delivered to opposing counsel or a regulator.
Technology-assisted review, where software helps prioritize and classify documents for review, is a well-established part of this category, but deploying it carries procedural prerequisites under the federal rules of civil procedure. The mechanics and the Rule 26 and Rule 502 requirements are covered in detail in How Technology-Assisted Review Actually Works. E-discovery platforms are usually the most expensive line in the stack, and the least forgiving of sloppy configuration.
What runs on the research and knowledge side?
Legal research platforms give lawyers access to case law, statutes, regulations, and secondary commentary. Citators — the tools that tell a lawyer whether a case is still good law — remain the discipline that decides whether research output is safe to rely on, a point examined in Verifying Authority in the AI Research Era: The Citator Discipline That Still Decides Cases.
Generative AI research assistants now sit in this category too, and they change the governance question rather than removing it. Courts have responded with standing orders on AI-assisted filings, and verification duties have not relaxed. Related reading: Judicial Standing Orders on Generative AI: The Filing Checklist Nobody Standardized. Some firms also run statutory-monitoring tools that track amendments across jurisdictions; see Knowledge Graphs for Statutory Monitoring.
Where do billing, payments, and firm operations sit?
Billing and time-recording software converts work into invoices, and payment tools collect on them. In firms serving regulated clients, this category overlaps with compliance: outside-counsel guidelines increasingly address AI-assisted drafting and what may be billed for it, a development tracked in Billing for AI-Assisted Drafting: How Outside-Counsel Guidelines Are Catching Up to Rule 1.5.
Around the core sits a connective layer: document management systems that store the firm's work product, e-signature tools, client portals, and secure file transfer. None of these is glamorous. All of them are where data lives, which makes them the surface a security review actually examines.
What does the AI layer change about the whole stack?
AI tools no longer occupy a single category. Meeting note-takers raise privilege questions; agentic tools that take actions in workflows raise supervision questions under professional-conduct rules; contract and research assistants raise verification duties. The governance issue is that these tools touch data and decisions across every category above, often procured without the diligence applied to the core stack.
Regulation is also arriving from outside the firm. State AI laws taking effect impose new documentation duties on legal and compliance teams, and the EU AI Act imposes obligations on vendors selling general-purpose models into Europe. Firms buying AI tools are increasingly the downstream party in those compliance chains. Practical steps follow. For related coverage, see EU AI Act Obligations for General-Purpose Models Took Effect August 2: What Legaltech Vendors Selling Into Europe Must Track.
What this means in practice
Three operational takeaways follow from the map above, each bounded by how the categories actually work.
- Map before you buy. Place every existing tool in a category, and identify overlaps and gaps. Most firms discover redundant subscriptions and at least one unowned category — often records retention or matter intake.
- Match diligence to data sensitivity. E-discovery and document management hold the most sensitive material, so vendor security review should be proportionate to what each category stores, not uniform across the stack.
- Govern the AI layer centrally. Because AI tools cut across categories, ad hoc departmental adoption creates gaps in privilege protection, verification, and supervision. A single intake process for AI tools is the practical fix.
The evidence for how firms run is mostly structural rather than statistical: the categories are stable, the AI layer is moving, and the professional-responsibility duties attached to each category are set by courts and regulators rather than by vendors. What remains unknown is how quickly the AI layer consolidates into the core categories — whether it becomes a feature of practice management and research platforms, or stays a separate procurement line. The stack map above will hold either way; the boundaries between boxes may not.
