// Independent data & AI consultancy · Manchester

Data first.
Everything else is
a tooling decision.

You're being asked for an AI plan while the market reinvents itself every six months. What doesn't change is this: businesses with solid data foundations absorb each wave and get stronger, while businesses that bolt AI onto a shaky estate pay for the same transformation twice. We help you establish a solid core before the money is spent.

Start a conversation The Readiness Report →

/ About Mimir

Mimir is an independent data & AI consultancy based in Manchester. The name belongs to the Norse keeper of the well of wisdom, and that's more or less the job: know what's true about your data before you build on top of it.

Tooling is a decision you should be able to remake cheaply; your data architecture is the thing you have to get right. Get the foundation sound and the market's constant reinvention starts working in your favour.

We're independent, not tied to any platform, vendor, or delivery practice, so the findings answer to your business rather than to what someone needs to sell you next.

We work best with leadership teams weighing a significant data or AI decision: organisations with a substantial data estate who want the case established before committing to a platform, a programme, or a partner.

// Based in Manchester, UK
// Established 2023
// Approach Data-first, independent
// Engagements Data & AI Strategy · Delivery · Speaking

Data first

Foundations before capabilities. The roadmap puts the structural work first, because everything interesting stands on it.

Independent

No reseller margins, no delivery arm waiting for phase two. If we recommend a tool, it's because it fits.

Traceable

Every recommendation points back to something found in your business during the process. None of it is lifted from a framework.

// Founder & Director

Robert Daly

Mimir is led by its founder, Robert Daly, a data and AI specialist with more than seven years spent building data functions, shipping data products, and turning advanced analytics into measurable commercial value. His work spans the NHS and digital health, financial services, legal, and manufacturing, from board-level strategy through to hands-on delivery in Python, Azure, and Databricks.

Robert developed Mimir's thirteen-domain assessment methodology, the engine behind the Readiness Report, and has designed data strategies and machine learning systems that remain in production use today. His research is published in BMJ Open, PLOS ONE, and JMIR, and he speaks regularly on responsible AI, data strategy, and AI readiness.

// Education MSc Neuroscience, King's College London
// Published in BMJ Open · PLOS ONE · JMIR
// LinkedIn Robert Daly

/ Services

Tools and models change by the month; your data estate is what they all rely on. Before you spend on AI, the Readiness Report establishes whether yours can support it.

Flagship · The Readiness Report

Data & AI Readiness Assessment

Every business creates valuable data through its daily work, and many are being pushed into AI before that data is in order. AI does not compensate for weak metrics, poor data quality, or low data literacy; it amplifies them. The assessment is built for the questions that carry real budget: what to fix first, what is safe to build now, and how to weigh what vendors are proposing.

Over three to four weeks we work inside your business, from your stakeholders' ambitions down to how your estate really operates. You receive a substantial written report: your maturity scored across the thirteen domains of Mimir's methodology, each with a defined level and a target to aim for; a prioritised roadmap that puts foundation work first; and a data flow map your whole team can use. Every recommendation is traceable to something found in your business, so the plan can be defended to a board, handed to a delivery team, or used to hold a vendor to account.

// ForLeadership teams weighing a serious data or AI spend
// DurationThree to four weeks, inside your business
// Scoring13 maturity domains, each with a defined level and a target
// DeliverablesWritten report · roadmap · data flow map · live readout
// CommercialsFixed fee, quoted after a short scoping call
// ObligationNone. The report stands on its own

01
Ambitions

Interviews with your senior stakeholders, and investors where relevant, on where the business is heading.
02
Estate

We map how data is captured, processed, stored and used, from ingestion through analytics, ML and AI.
03
Assessment

Your maturity scored across thirteen domains, each with a defined level: a line in the sand today, a benchmark for tomorrow.
04
Roadmap

A prioritised set of recommendations, and the order to work through them.
05
Readout

A modern data flow map and the full report, presented back to your team by ours.

Commission a Readiness Report Download the one-pager →

Delivery Support

The report is written so your own team can deliver the roadmap. Where clients want continued support, we stay involved directly or introduce trusted specialists in areas like data engineering and hardware. If you'd rather appoint your own delivery partner, we help you choose well; if you need to hire, we bring in recruitment partners. In either case the roadmap, and the independence it was built on, remain yours.

Hands-on or advisory
Specialist & recruitment partners
Vendor selection & oversight

Speaking & Panels

We talk about data strategy, AI readiness, and what goes wrong in between, on stages and around tables. If you're putting an event together, we'd like to hear about it.

Talks & keynotes
Panels & round tables
Workshops & private briefings

/ Track record

Engagements are confidential by default, so names are withheld here. The work is real, recent, and measured; ask about any of it on a call.

// National health body

Machine learning in production

Embedded with the education arm of a national health service across a portfolio of data science work: a team-wide data strategy, a masterclass series that lifted the team's data capability, and an anomaly detection system that identifies misuse of a national learning platform.

Anomaly detection live in production since 2023
Data strategy adopted across the team

// Scale-up with a multi-system estate

From readiness to a unified platform

A thirteen-domain readiness assessment gave the leadership team a prioritised data and AI roadmap. The first foundation item followed immediately: a lakehouse platform unifying four separate source systems into one governed analytics estate.

Four source systems unified into one platform
Over a day of reporting effort saved every week

// Established technology consultancy

Positioning an AI practice

Sharpened the AI service offering of a technology consultancy in a crowded market, grounded in competitive analysis of AI adoption across its target sectors: where to accelerate, what to build, and how to differentiate.

AI service portfolio refined and repositioned
Sector-by-sector adoption analysis

// World-leading university

Evaluating innovation with network science

Applied social network analysis to evaluate health technology accelerator programmes, giving programme leads evidence of how their network actually behaves rather than how it is assumed to.

Programme evaluation grounded in network data
Findings delivered to programme leadership

// Digital health scale-up

Building a data function from zero

Built and led the data function of a digital health scale-up over four years: a data team grown from scratch, fifteen-plus colleagues trained in data literacy, and data turned from an internal cost into a revenue line through new products and enterprise channels.

£180k Innovate UK grant won and delivered as an NLP model
Data-as-a-service product onboarding four major pharmaceutical companies

// Mental health technology startup

Responsible AI where it matters most

Designed and built an LLM-powered mental health agent for young people, with safety guardrails and safeguarding principles applied throughout, alongside a mood-detection model built on passive phone-sensor data and taken from concept to MVP.

LLM agent for a safeguarding-critical audience
Sensor-based mood detection, concept to MVP

/ Contact

Tell us what you're working on and where it's stuck; the short version is fine. Messages come directly to us and we reply within two working days. The first step is a short call to establish whether we can help, with no obligation on either side. For speaking and panels, please include the date and the format.

// Location Manchester, United Kingdom
// Response time Within 2 working days
// First step A short introductory call
// LinkedIn Mimir Data Ltd

Data first. Everything else is a tooling decision.