AI for Professional
Services Firms
Professional firms need to reduce unbilled work without weakening judgement, confidentiality or review. We build AI into defined workflows with source evidence, matter or client permissions and a named professional owner.
Search the firm knowledge
Keep client permissions
Show the source
Staff find useful past work without opening every file.
Sourced
OUTPUTS LINKED TO APPROVED RECORDS
Reviewed
PROFESSIONAL APPROVAL BEFORE USE
9 Mar 2026
RICS REQUIREMENT FOR IN-SCOPE SURVEYING
AI needs to fit your professional duties
AI tools can help prepare work, but the practice remains responsible for confidentiality, accuracy and supervision. The workflow needs to reflect the duties of the relevant profession.
Specialist platforms such as Harvey and CoCounsel can be a good fit, but enterprise buying, per-seat pricing and minimum commitments do not suit every practice. We compare them with a narrower custom workflow before recommending a build.
We build private workflows around the work and controls of your practice.
WHAT MOST FIRMS HAVE
- A ChatGPT tab on each partner's laptop
- Client data going into tools nobody signed an agreement with
- No audit trail, no log, no policy
- Knowledge still trapped in three partners' heads
- A product that misses a required workflow
WHAT YOU NEED
- Workflows wired into your matter system, not bookmarks
- Models that don't train on your client data, contractually
- Every output sourced, logged, reviewable
- Your firm's precedents and house style included
- Commercial scope agreed before systems are connected
Professional-services work worth automating
The system names differ between law firms, accountancy practices, consultancies, surveyors and recruiters. These are usually the most practical places to start.
Intake and conflicts
Collect enquiry details, organise evidence for conflict and KYC checks, and prepare an engagement draft. Authorised staff review the checks before accepting the client or opening the matter.
First-pass drafting
Prepare drafts from approved precedents and source documents. The responsible professional checks the facts, reasoning and suitability for the client.
Document review
Organise and summarise bundles with page references for review. The pilot measures missed issues and false flags as well as preparation time.
Time capture and WIP
Prepare draft time entries from approved work records. Staff check the activity, duration and matter before billing. Monitoring boundaries and privacy requirements are agreed first.
The firm's brain
Make approved precedents, manuals and recorded practice knowledge searchable. Experienced staff help resolve gaps, while permissions prevent access across restricted clients or matters.
How we start
We start with one professional workflow, representative client work and a defined review standard. The evidence from that phase determines whether a wider programme is justified.
Our relevant work covers recruitment, business-finance compliance, financial advice and human-reviewed assessment. Each practice still needs its own professional rules, data boundary and review owner.
BOOK A PRACTICE AUDITPractice audit
We work with partners, senior staff, operations and technology to map the work and its professional controls. Candidate workflows are ranked by value, readiness and the consequence of a wrong output.
Capture the firm's knowledge
Precedents, advice files, model letters, internal manuals, the bits in partners' heads. Pulled into a private knowledge layer your workflows can read. Hosted on infrastructure you can point your regulator at.
Launch the first workflow
Test a contained workflow with representative client work, source references and a professional approval step. The existing process remains in control until the pilot is accepted.
Add the rest, sensibly
Next workflow, then the one after. Each one comes with the regulatory paperwork it needs: risk register, client notice where appropriate, supervision policy and model card. We reuse the common parts rather than starting again.
Professional duties still apply to AI work
Professional standards continue to apply when AI assists the work. The requirements below have different scopes, so we check which affect your practice.
Ayinde v Haringey.
The June 2025 judgment addressed non-existent authorities and referrals to professional regulators. Lawyers must verify sources and the propositions they support before relying on AI-assisted work.
AI requirements for RICS surveying practice.
RICS's responsible-use standard applies to RICS members and regulated firms in surveying practice. The controls depend on how AI affects the professional service, with professional judgement and oversight retained where the impact is material.
Guidance on generative and agentic AI in audit
The FRC's March 2026 guidance discusses audit-quality risks, possible mitigations and professional judgement when assessing AI outputs. It provides guidance and illustrative examples, rather than a new universal set of AI audit rules.
Sources: Ayinde and Al-Haroun judgment, RICS responsible-use standard, FRC AI in Audit guidance.
Our experience includes accountancy and regulated finance
Alex built AccountancyManager through to its sale to Hg-backed Bright. Crystal automates compliance audits for business finance, while RiskMapped is being developed around the protection-analysis work of financial advisers.
When this is worth discussing
We work best when there is a real operating problem, enough volume to measure and people from the affected teams who can make decisions.
Usually a good fit
- An established UK business, usually with annual revenue above £10m
- A repeated process with a known cost, delay, error rate or capacity problem
- A senior sponsor and a day-to-day owner who understand the work
- Access to the relevant staff, systems, sample records and security requirements
We may point you elsewhere
- A standard product already covers the process well
- The requirement is a one-off small build with no wider operating case
- There is no owner or access to the people and data needed to test the result
- The plan relies on AI making high-impact decisions with nobody responsible for review
Questions the buying team will ask
If AI does the first draft, what stops our juniors learning?
The firm needs an explicit training plan. Junior staff should still practise the underlying work, check sources and receive feedback on their reasoning. We agree how the workflow supports that development with the supervising professionals.
Our client data can't leave the building. How does that work?
The provider, product tier and configuration determine the data terms. We review training use, retention, subprocessors, residency, access and logging against your contracts before selecting a model or hosting option. The resulting data flow and responsibilities are documented for your privacy and security teams.
We looked at Harvey / CoCounsel. Why you instead?
An established legal AI product may be the right answer when its features, data terms and integration fit the practice. Custom work is justified where a defined workflow, precedent set or connected system creates a measurable requirement the product cannot meet. We compare those options during scoping.
If AI drafts in minutes, doesn't our hourly billing model break?
It may change the economics of some work. Finance and the practice owners need to assess the effect on capacity, fees and client expectations. The business case should use actual effort and review costs rather than assume every saved hour becomes profit.
What if the AI hallucinates a case or a clause?
It can. The workflow should retrieve from approved precedents, statutes, contracts or matter documents and cite the relevant source for review. The responsible professional still checks the conclusion before it is used.
How long, and how much?
The first phase covers one defined workflow, representative client or matter data, professional review and a value baseline. We scope the systems and controls before issuing a commercial proposal.
Who owns what you build?
The commercial proposal states ownership of source code, prompts, knowledge sources and audit records, together with hosting, support and handover terms. We can work in your cloud account or operate an agreed environment.
Talk to us about your firm
Tell us which work absorbs senior time, the systems involved and the professional duty that applies. We will identify a first workflow that creates value without transferring judgement to the model.