Professional Services
Make what the firm already knows findable, safely.
Governed enterprise search, document intelligence and secure copilots that respect matter confidentiality and conflict boundaries.
Operational context
How the sector actually runs.
Expertise is expensive and unevenly distributed. A firm's best answer to a client question usually already exists in a previous engagement, but finding it depends on knowing who did that work and whether they are still here.
Confidentiality boundaries are not optional. Matter and engagement separation, conflict walls and client-imposed restrictions mean a search tool that ignores permissions is not a productivity feature, it is a professional liability.
Revenue depends on timesheets nobody wants to complete. Unrecorded time, late entry and write-offs erode realisation, and the people best placed to record time accurately are the ones least inclined to stop and do it.
Onboarding is repetitive and rules-heavy. Engagement letters, conflict checks, identity verification, matter setup and system provisioning follow the same sequence every time, with different thresholds by service line and jurisdiction.
Precedent documents are copied and quietly diverge. A template is duplicated for a client, amended, and becomes the new de facto precedent for whoever copies it next, so the firm's standard position drifts without anyone deciding it should.
High-value problems
Where the money and the risk actually sit.
We have done this work before but I cannot find who did it
A permission-aware search layer over matter files, engagement documents and work product returns cited answers restricted to what the person asking is already entitled to see. Conflict and ethical walls are enforced at retrieval, not filtered afterwards.
I am not letting AI near client documents
That is the correct instinct, and it is the design constraint we start from. Retrieval is scoped to the user's existing entitlements, prompts and outputs are logged, data is not used to train a vendor model, and the whole flow is reviewable before anyone turns it on for a client matter.
Our realisation is dropping and we find out at month end
We build reporting that connects recorded time, billed value and write-offs at engagement level, refreshed daily, so a partner sees a realisation problem while there is still something to do about it.
Onboarding a new client takes three days of somebody's life
Conflict checks, identity verification steps, engagement letter generation and system provisioning are sequenced into one workflow with the thresholds encoded per service line. A person still approves the engagement, but they approve a completed pack rather than assembling it.
Everybody uses a slightly different version of our standard agreement
Document intelligence compares executed documents against the current precedent, surfacing where standard positions have drifted and how often. The firm decides which drifts are improvements and which need pulling back.
Evidence from the field
Where this has actually worked, and where it has not.
Published results from other organisations, cited so you can check them. None of these are our clients and none of these numbers are ours.
Where the gains are real
This is the sector with genuine independent benchmarking, and the results are stronger than most partners expect. On defined document tasks, current tools now beat a lawyer baseline. On tasks needing judgement across a whole matter, they do not.
What the evidence does not support
The benchmark that produced these numbers was declined by three of the largest legal research vendors, so it measures the tools that opted in. Separately, none of this evidence addresses privilege, conflicts or confidentiality, which is where firm-level deployments actually stall.
Vals Legal AI Report, with a human lawyer control group
Global, using real law firm data
The first independent benchmark of legal AI tools across seven tasks lawyers actually do, including data extraction, document question answering, summarisation, redlining and chronology generation.
AI outperformed the lawyer baseline in four of the seven areas tested. On document question answering the best tool scored 94.8 per cent against a lawyer baseline of 70.1 per cent.
What it does not prove: Thomson Reuters, LexisNexis and vLex did not participate. The lawyer baseline is a control group, not a firm's best available specialist.
Vals Legal AI Report, legal research extension
Global
Extended the benchmark to legal research specifically, comparing three legal AI systems and one general foundation model against a lawyer baseline.
All AI systems landed around 80 per cent accuracy against a lawyer baseline of 71 per cent.
What it does not prove: 80 per cent accuracy means one answer in five is wrong. That is a review workflow, not an answer service.
Noy and Zhang, published in Science
Peer-reviewed research
A randomised experiment on mid-level professional writing tasks of the kind consultants, marketers and analysts do daily.
Time taken fell about 40 per cent and assessed output quality rose about 18 per cent, with the largest gains going to the lower-performing participants.
What it does not prove: Measured on discrete 20 to 30 minute tasks. A task speed-up is not the same as a billable-hour or realisation improvement.
Brynjolfsson, Li and Raymond study of a Fortune 500 software firm's customer support operation
United States (peer-reviewed working paper)
Staggered rollout of a generative AI conversational assistant to customer support agents, with agents who had not yet received it acting as the comparison group.
Across 5,179 customer support agents, access to the assistant raised productivity, measured as issues resolved per hour, by 14 per cent on average. Novice and low-skilled workers improved by 34 per cent, while experienced and highly skilled workers saw minimal effect.
What it does not prove: One firm, one job type, and an early-generation assistant, so the size of the gain is specific to that setting. The paper was revised before journal publication in the Quarterly Journal of Economics, so figures differ slightly between versions and you should quote the version you link to.
Ashurst (now Ashurst Perkins Coie)
Global law firm, 23 offices across 14 countries
Firm wide trials of three generative AI tools across legal and business services tasks, including a blind study in which an expert panel scored AI drafts against lawyer drafts without knowing which was which.
411 partners, lawyers and staff took part between November 2023 and March 2024. Reported time savings of roughly 80 per cent on drafting UK corporate filings, 59 per cent on industry research reports and 45 per cent on first drafts of legal briefings. In the blind study, AI outputs averaged 3.0 out of 5 for accuracy against 3.5 for lawyer written outputs, and the panel correctly identified every human written output as human.
What it does not prove: Published by the firm itself and not independently audited. The savings are task level, measured on set exercises rather than on billed client matters, and the same report shows AI accuracy scoring below the firm's own lawyers.
METR (Model Evaluation and Threat Research)
Randomised controlled trial on mature open source repositories
Experienced open source maintainers were randomly permitted or not permitted to use AI tools on real issues in repositories they already knew well, with completion times recorded.
Across 16 developers and 246 issues, developers took 19 per cent longer to complete work when AI tools were allowed. The same developers had expected AI to speed them up by 24 per cent beforehand, and still believed they had been 20 per cent faster afterwards.
What it does not prove: Small sample in a narrow setting: expert maintainers on codebases they know deeply. METR explicitly states the result is not evidence that AI fails to speed up most developers, and that it reflects the tools available in early 2025. The value here is the gap between perceived and measured speed, not the 19 per cent itself.
Linklaters
United Kingdom, English law
The LinksAI English law benchmark: 50 questions across 10 practice areas, marked out of 10 (5 for substance, 3 for citations, 2 for clarity) against the standard of a competent mid-level lawyer with two years post qualification experience.
In the February 2025 round, OpenAI o1 scored 6.4 out of 10 and Gemini 2.0 scored 6.0, against a best score of 4.4 out of 10 in the October 2023 round. Linklaters concludes the models should not be used for English law advice without expert human supervision, though they may be useful for a first draft or a cross-check in well-known areas of law.
What it does not prove: A 50 question benchmark designed and marked by one firm, not a measure of work delivered to clients. Model versions change quickly, so the scores date fast.
Deloitte Australia and the Department of Employment and Workplace Relations
Australia, Commonwealth government
An independent assurance review of the Targeted Compliance Framework, in which a generative AI tool chain was used to assess whether system code could be traced to business requirements.
The report published on 3 October 2025 was withdrawn after fabricated academic references and an invented quotation attributed to a Federal Court judgment were identified. The revised report discloses the use of "a generative AI large language model (Azure OpenAI GPT 4o) based tool chain" licensed by the department. Deloitte agreed to repay the final instalment of a contract worth just under A$440,000. The department published a further corrected version on 3 February 2026.
What it does not prove: The department's page records only that corrections were made and that the report replaces the version published on 3 October 2025. The detail of the fabricated citations and the repayment comes from reporting and from Deloitte's own statements, not from a published audit, and the refund amount was not disclosed at the time.
Sector figures are useful for deciding what to try first, and almost useless for forecasting your own result. The organisations above had the data, the systems access and the governance in place before they measured anything. The full global picture covers 13 industries, sets out what separated the organisations that got a result from the ones that got nothing, and carries the randomised trials that found no gain at all.
Use cases
What we build in this sector.
Each is tagged with the disciplines involved, because most useful work crosses more than one.
Governed enterprise search
Retrieval across matter files and work product with entitlements and ethical walls enforced at query time, and every answer cited to its source.
- AI
- Security
Secure client-facing copilots
Assistants scoped to a single engagement's documents, with logged prompts, reviewable outputs and no vendor training on firm data.
- AI
- Security
Document and clause intelligence
Extract obligations, dates and clause positions from executed documents, and compare them against the firm's current precedent.
- AI
- Data
Client onboarding workflow
Sequence conflict checks, verification, engagement letters and provisioning into one auditable flow with per-service-line rules.
- Automation
- Security
Realisation and utilisation analytics
Connect recorded time, billed value and write-offs at engagement level with a daily refresh rather than a month-end reveal.
- Data
Proposal and pitch assembly
Draft proposal sections from prior approved material, with provenance shown so a partner can see what each paragraph was taken from.
- AI
- Automation
Identity and access governance
Access reviews, joiner-mover-leaver automation and privileged access control appropriate to a firm holding confidential client material.
- Security
- Cloud
Human oversight
Where a person still decides.
Automation proposes. A named person approves. These are the points we design the workflow to stop at.
- A partner or responsible principal approves any AI-assisted work product before it reaches a client. Nothing generated is sent without professional review.
- The conflicts team approves every new engagement following the automated check. An automated clear result is a prompt for review, not a decision.
- A responsible partner approves enabling any copilot on a specific matter, with the entitlement scope recorded.
- The practice manager approves any change to entitlement rules or ethical wall configuration, separately from IT change control.
- A billing partner approves write-offs and adjustments surfaced by realisation reporting.
Delivery options
On Azure, on AWS, and afterwards.
The platform is chosen for the workload, not for us. Both paths end in the same place: someone still owns it after go-live.
- 01
Delivered on Microsoft Azure
The natural fit for firms already on Microsoft 365, where matter documents live in SharePoint and confidentiality is enforced through existing site and label permissions.
- Azure AI Search with security trimming inherits SharePoint and Entra ID permissions so retrieval cannot cross an ethical wall.
- Microsoft Purview applies sensitivity labels and data loss prevention to client material, including in AI interactions.
- Microsoft Foundry hosts the retrieval and drafting workflows with prompt and output logging retained for review.
- Microsoft 365 Copilot and Copilot Studio deliver assistants where fee earners already work, subject to the firm's approval process.
- Microsoft Entra ID Governance runs access reviews and joiner-mover-leaver automation.
- Microsoft Fabric and Power BI deliver realisation and utilisation reporting with partner-level row security.
- 02
Delivered on AWS
Suits firms with a bespoke practice management platform, an existing AWS estate, or a requirement to keep the knowledge layer independent of the productivity suite.
- Amazon OpenSearch Service provides retrieval with document-level access control mapped to matter entitlements.
- Amazon Bedrock runs retrieval and drafting, with Bedrock Guardrails restricting responses to retrieved firm content.
- Amazon Q Business delivers governed search over connected firm repositories.
- AWS IAM Identity Center centralises access, with Amazon Macie identifying sensitive client data at rest.
- AWS Step Functions orchestrates the onboarding workflow with a complete audit trail per engagement.
- Amazon Redshift and Amazon Quick deliver realisation and utilisation reporting.
- 03
Operated after handover
Permissions change constantly as people move between matters. A retrieval system is only as trustworthy as the last time its entitlements were verified.
- Entitlement synchronisation verified continuously, with any retrieval crossing an ethical wall treated as a priority incident.
- Answer quality and citation accuracy sampled and reviewed with the knowledge team.
- Prompt and output logs retained per the firm's records policy and available for professional review.
- Access reviews run on the agreed cycle, with exceptions escalated to the practice manager.
- Model and platform changes assessed before adoption, including any change to vendor data handling terms.
- Monthly reporting on usage, escalation rates and where the knowledge base is failing to answer.
Related services
The capabilities behind this work.
Artificial Intelligence
AI that is chosen for a reason, costed before it is built and governed once it is live.
Security and Governance
Close the ways in, know exactly who can do what, and answer an auditor or a client questionnaire from current evidence rather than memory.
Data and Analytics
Reporting your executives trust, produced once, with a modelled layer underneath it that the next AI or forecasting project can stand on.
Managed Services
Your platform keeps earning its business case after go-live, with cost, security posture, reliability and adoption reviewed on an agreed cycle rather than left to drift.
Free discovery workshop
Start with one professional services challenge.
Bring a process that costs more than it should. We will map the opportunity, the readiness gaps and a recommended next step.