Services

Three capabilities.
One way of working.

Each capability solves a different problem. Most clients begin with the AI Operating Audit, then commission one or more of these. The retainer that follows is what keeps the systems performing and up to date after launch.

4 weeks Fixed price The starting point for every engagement

The AI Operating Audit

Before we build anything, we map what you actually need.

The Audit is a structured diagnostic that examines your processes, your document flows, your existing tools, and the work currently consuming senior time. We interview key staff, review historical outputs, identify the highest-ROI use cases, and deliver a prioritised twelve-month roadmap with clear sequencing, scope, and budget estimates for each phase. Think of this as Business Process Engineering.

Who it's for

Decision-makers who know AI matters but don't yet have a defensible plan.

What you get

/01 An inventory of AI-suitable workflows ranked by ROI and feasibility
/02 A data and document readiness assessment, with specific gaps identified
/03 A risk and governance review covering EU AI Act, GDPR, sector-specific regulation
/04 A twelve-month phased implementation roadmap
/05 A board-ready summary document and presentation

Request audit scope and pricing

2–4 months Engagement From €25,000

Agentic Workflows

We automate the multi-step processes that consume your team's time.

Most "AI agent" pitches you've heard are either hype or repackaged automation. We're direct about what these systems can and can't do.

We build AI-augmented workflows that handle structured, repetitive, multi-step work — with human oversight where judgement matters. Every workflow we build has explicit human-in-the-loop gates, full audit trails, and an evaluation framework so you can measure performance over time. Nothing autonomous goes live without controls.

Use cases

/01 Procurement and tender support — submission checking, gap analysis, compliance verification
/02 Regulatory compliance workflows — directive checking, policy alignment, audit preparation
/03 Document intake and routing — classification, extraction, workflow assignment
/04 Reporting cycles — automated draft generation from structured data, with review built in

Discuss a workflow

3–4 months Engagement

Knowledge Retrieval

the the Knowledge Engine

Your accumulated knowledge, instantly searchable, with sources you can verify.

Every organisation we've worked with has the same problem: somewhere in your SharePoint, Google Drive, email archive, or someone's hard drive sits the answer to most of the questions your team asks every day. Finding it takes hours — when it gets found at all.

We build private, secure Knowledge Engines that index your historical content, then let your team ask natural-language questions and get cited, source-grounded answers. This is the same foundational architecture we built for ligi.ai — adapted to your private corpus, your security model, and your team's actual workflow.

Built around

/01 Citations are mandatory — every answer points back to source documents. No black-box outputs.
/02 Access control is preserved — the system respects your existing permissions matrix.
/03 Hosted in the EU — your data does not leave compliant infrastructure.
/04 Evaluation framework included — measurable accuracy and quality from day one.

Explore a Knowledge Engine

2–3 months Engagement

Document Generation

the Drafting Acceleration

First-draft proposals, reports, and regulated documents in your firm's voice.

The slowest part of professional services work is producing the first draft of a long, structured document. Senior people spend hours on the parts that don't require senior judgement, leaving less time for the parts that do.

We build Drafting Acceleration systems that turn your historical work into a structured drafting assistant. The system produces a 60–70% complete first draft in your firm's voice. Your senior people focus on the strategic decisions, the bespoke analysis, and the final review — the parts where your value actually sits.

What we do differently

/01 We tune to your voice — drafts read like your firm wrote them, not like a generic LLM
/02 We design for review, not replacement — human gates are built in, not bolted on
/03 We protect institutional learning — junior staff still learn the craft
/04 Often paired with a Knowledge Engine, since the two compound

Discuss your drafting workflow

Ongoing retainer

Managed Operations

The retainer that keeps your systems performing.

Production AI systems are not "set and forget." Models change. Data evolves. Workflows shift. Without ongoing monitoring and tuning, performance degrades — and the most common reason AI projects fail to deliver is that nobody owned them after launch.

Every system we build is supported by a Managed Operations retainer. This is not a maintenance contract; it is the team that keeps your systems improving over time.

Includes

/01 Performance monitoring and ongoing evaluation
/02 Prompt and configuration tuning as your needs evolve
/03 Onboarding of new document types, users, or use cases
/04 Quarterly review and roadmap updates
/05 Direct access to your delivery team

Discuss ongoing support

8 weeks Fixed price For organisations on the Microsoft stack

Microsoft AI Activation

You've paid for the licences. Now make them work.

Most organisations buy Microsoft 365 Copilot, watch a few demos, roll it out to staff, and then watch adoption stall. The licences become a line item nobody owns.

Microsoft AI Activation is the eight-week engagement that turns your existing Microsoft estate into a working AI operating layer. Not training sessions. Not generic prompt libraries. Not "Copilot champions" programmes. Real, measurable workflows built on Copilot, Copilot Studio, Azure OpenAI, and Power Platform — designed for the specific work your organisation actually does.

Who it's for

Organisations already invested in Microsoft 365 — Copilot licences live, SharePoint estate in place, Power Platform available — that need to actually realise the value of what they've paid for.

What you get

/01 A Copilot adoption audit identifying where the licences are working and where they aren't
/02 Custom Copilot Studio agents tied to your specific work — tender review, contract checking, compliance Q&A, drafting assistance
/03 Knowledge integration connecting Copilot to your SharePoint and document libraries with proper grounding, permissions, and citation behaviour
/04 Power Platform automations that turn conversational AI into operational capacity
/05 An adoption framework, governance model, and 12-month roadmap

Explore Microsoft AI Activation

How we work

Every engagement, four stages.

Stage 01

Diagnose

Start with the Audit or a scoped equivalent. Nothing gets built without a clear understanding of the problem.

Stage 02

Design

Document the workflow, the data flows, the user experience, and the governance controls before any code is written.

Stage 03

Build & Pilot

Build in short iterations with the eventual users involved throughout. Pilot before production.

Stage 04

Operate

Hand over with full documentation and a Managed Operations agreement that keeps the system performing.

Day one starts

Start a project

Book a discovery call and we'll diagnose the problem first.

Three capabilities. One way of working.