In practice

The work, not the pitch.

Three anonymised engagements — different sectors, different problems, the same approach. Where work is still in progress, we say so.

Building the intelligence layer a growing restaurant group didn't have

The challenge

A multi-site hospitality group operating across two brands and seven locations had no consistent, real-time view of how the business was performing. Management information arrived at month-end — long after the decisions it should have informed had already been made. Labour cost, the single largest variable in the P&L, was tracked by approximation. On any given morning, the operator had no reliable way to know whether the business had a good day or a bad day yesterday.

The business had a clear ambition — a single version of the truth, with every site aligned around the same metrics, using up-to-date data that is clear, consistent, and drives action. It had no infrastructure to deliver it.

The work

Tenex designed and built an automated MI pipeline from the ground up. Data flows from the point-of-sale system and workforce management platform into a central reporting layer every morning before 8am. Site managers receive a daily flash — sales against prior year and the prior week, labour cost, labour percentage and sales per labour hour, colour-coded by threshold — before the working day begins.

A weekly trading summary and a period-level P&L, with a dynamic labour target that adjusts to actual sales performance rather than a fixed budget, followed in subsequent development phases. The pipeline handles two different brand configurations, planned closures and the structural differences in how each brand's data is reported.

Five KPIs were agreed and embedded across the organisation: like-for-like sales growth, transaction volume, average transaction value, gross profit percentage (target ≥70%) and sales per labour hour. These are not metrics management tracks periodically — they are the daily language of the business.

Where it stands

Daily MI is live across all seven sites. The management team receives automated same-morning trading data it did not have before. Weekly trading and period P&L are in the build phase.

The business owns the infrastructure. When the engagement moves on, the pipeline stays — along with the capability to read and act on it every morning.

7 Trading sites across two brands — all covered by a single automated pipeline
5 Company-wide KPIs the whole business is now aligned around
08:00 Daily trading flash delivered automatically — before the working day starts

Embedding AI into a specialist bid process — without losing what makes it win

The challenge

A specialist bid consultancy in the facilities management and construction sector had deep expertise in winning complex public sector contracts. The problem: that expertise lived almost entirely in individual team members. Each new bid required the same knowledge to be rediscovered from scratch. With government procurement frameworks becoming more demanding, and clients increasingly expecting AI-assisted capability, the need to systematise was clear.

The team had access to Microsoft Copilot. What it lacked was the architecture, governance and prompting discipline to use it reliably on live bids — without introducing risk to the quality or compliance of the output.

The work

Tenex mapped the existing 22-step APMP-validated bid process and identified the highest-value AI intervention points: answer planning, evidence retrieval, content drafting and consistency checking. A two-stage prompting architecture was built and tested — document discovery first, then verbatim content retrieval with mandatory source citation — removing the risk of the AI conflating sources or producing unverifiable assertions in bid responses.

An AI governance framework was developed and issued, compliant with government AI transparency requirements. Copilot agents were configured and scoped to individual client knowledge bases within the firm's Microsoft 365 environment, with IP separation designed in from the outset.

The first major client — a tier one construction contractor — is actively adopting the capability, moving toward AI-assisted bid responses using their own content library and the implementation architecture developed jointly with Tenex.

Where it stands

The AI-enabled bid operating system is live and in use on current bids. Efficiency data from the first live deployment: a confirmed 20% time saving on a complex government framework bid — ten days budgeted, eight days actual. Validated by the proposals director and now used as the basis for pricing AI-assisted bid work.

The consultancy can now offer clients AI-assisted bid development as a product — not as an experiment, but as a governed, evidenced service. The capability sits inside the business.

20% Time saving confirmed on a live government framework bid — validated and used as future pricing evidence
22 Step APMP-validated bid process mapped, with AI intervention points identified and built
7 Government framework lot types covered by the structured bid knowledge base

Laying the infrastructure for a business built to grow — and eventually sell

The challenge

A UK food manufacturer with a strong product and established retail relationships was running ahead of its operational infrastructure. A new production facility was being planned with no digital tool to model layout options. Employment compliance was managed reactively. Management information was thin. And the founder, who had built everything himself, was making most decisions in isolation — without the governance structures that would allow the business, or its value, to survive his absence.

A significant new export order had just been secured. The business needed to scale quickly and cleanly — but without the systems in place, the compliance and commercial exposure would grow faster than the opportunity.

The work

Three workstreams running in parallel alongside strategic and financial advisory involvement.

First, an employer compliance platform — a structured operational reference covering 124 compliance items across employment law, health and safety (including COSHH, food production machinery, manual handling and fire safety), training obligations and HR situations. Built as a working tool for the production management team, not a policy document.

Second, a 3D digital model of the new production facility — built to the exact dimensions of the space, including structural pillars, machinery footprints and circulation routes. Used to plan the fit-out before any physical layout decision is committed, and as a communication asset for future investor conversations.

Third, financial analysis, growth strategy and commercial structuring — including a supply agreement for a new major customer relationship covering order volumes, cancellation protection and ingredient procurement liability; alongside ongoing work on management capability, governance visibility and delegated decision-making as the building blocks of a credible exit story.

Where it stands

The compliance platform is live and in use. The 3D facility model is complete and informing layout decisions. The commercial and governance infrastructure is being built in parallel with the operational scale-up.

The goal is a business that runs without depending on its founder — one a buyer would pay a premium for. That work starts now, not in the final year before sale.

124 Compliance items across employment law, H&S and training — operational reference, not a policy document
3D Digital facility model built to exact specification — used for layout planning and investor conversations
3–5yr Exit horizon — governance and management infrastructure being built toward a structured sale

Recognise any of these problems in your business?

The first conversation is about your business — what you are dealing with, where the friction is, and whether Tenex is the right fit. No pitch. No obligation.

Get in touch