Data transformation and bespoke models

Keep the systems you have, and make the data work harder.

If you are happy with what you have but need someone to build the Power BI reports, transform the data, or create a model that works the way the HR Intelligence engine does, I build it.

I am not a coder by background. My architectural experience means I understand what the end user needs and how to put it together, and I use modern AI-assisted development to build it properly, then stress test it.

What I build

Reports, transformations and models.

  1. Power BI reports

    Reports and models built on a clean star schema, with the measures documented and the numbers matching their source, designed so a busy manager sees the point in seconds.

  2. Data transformation

    Exports from different systems and spreadsheets turned into one clean, consistent model, the same way every month, with every step visible.

  3. Bespoke models

    A model built around one team’s job, where everyone works in their own copy and it merges into a master that does the MI automatically.

  4. Spreadsheet models made fit

    From making end-user computing models fit for purpose, to engineering bespoke data tools. You get a working tool your team can own, with its logic documented and tested.

Bespoke models

From each person’s own tracker to one master and its MI.

Employee relations model

Who uses it
Each ER adviser, the ER lead and the HR Director.
What it tracks
Each adviser works in their own tracker and uploads it into one master, which merges every adviser’s grievances, disciplinaries, absence cases and data breaches, so nobody is locked out of a shared spreadsheet.
The MI it gives
The MI produces itself: caseload by adviser and type, days open and time to resolve, ACAS steps recorded, tribunal-risk prompts, FCA conduct flags and manager hot spots.

Recruitment engine

Who uses it
Recruiters, hiring managers and the Head of Recruitment: a team that wants a tracker it can work with rather than a recruitment HRIS.
What it tracks
Job requisitions and openings, every application from applied to hired, sources, recruiters, offers and starts.
The MI it gives
Time to fill and to accept, drop-off by stage, offer acceptance, recruiter and source performance, early leavers by source and the cost of open vacancies, on one dashboard.

Learning dashboard

Who uses it
Each learner, their manager, and the learning and compliance team.
What it tracks
Mandatory modules with due and expiry dates, completions, training delivered and the educators who deliver it.
The MI it gives
Each learner’s own list of what is due, overdue and done; compliance by team and division; what expires next; educator KPIs.

Survey model

Who uses it
HR, People and Culture, and the owners of onboarding and exit.
What it tracks
Questionnaires built for you (new starter, exit, pulse), who is due from the headcount dates, and the send lists.
The MI it gives
Results by theme and statement, a recommend score with its base, a word picture that never keeps comments, and small groups protected by a threshold your organisation sets.

And across the lifecycle

  1. Absence case management

    Open and long-term absences, return to work, Bradford triggers and the actions due on each case.

  2. Year-end reward and calibration

    Ratings round trip with managers, pay review and bonus modelling, compa-ratio and pay equity checks.

  3. Onboarding and probation

    New starters, probation reviews due and overdue, new-starter risk and the six-week check.

  4. Leavers and exit

    Future leavers, the team each one leaves behind, regretted loss and exit themes.

  5. Establishment and organisation change

    Headcount against budget and establishment, approved changes, and HR and Finance reconciled.

  6. Data room packs

    Headcount, liabilities and contracts reconciled and ready for buyer scrutiny in a sale or merger.

  7. Statutory returns

    Pay gap and other returns prepared and checked back to source.

  8. Any team, any tracker

    Wherever a team keeps its work in spreadsheets that someone has to merge by hand, the same pattern applies.

Spreadsheet models made fit

End-user computing, brought under control.

Regulated firms run important figures through spreadsheets built outside formal IT controls. Making them fit for purpose means the same disciplines a system would have.

  1. An inventory

    Every model that matters found and listed, with what it feeds and who relies on it.

  2. Risk tiers

    Each model graded by its impact and complexity, so the critical ones get the strongest controls.

  3. Owners and documentation

    A named owner, a user guide, the logic written down and a log of every change.

  4. Version history and access

    Who can change it, and every version kept.

  5. Tested

    Checks written to fail if something breaks, and the rebuilt model’s outputs proved against the original before it is switched over.

  6. Rebuilt where it should be

    Where a spreadsheet has outgrown itself, the same logic rebuilt as a model or a report, with nothing lost.

How I build

Built from parts already proven in the engine.

Each model is shaped around one team’s job from parts that already work in the HR Intelligence engine, and delivered in the browser, in your spreadsheet, or hosted for a team with who sees which case agreed with you first.

  1. Upload any export

    CSV or Excel, one file or a folder of months.

  2. Mapping desk

    Unknown columns decided once, then remembered.

  3. Audit before load

    What is broken, ranked by what it affects.

  4. One model

    Many systems and spreadsheets read as one.

  5. KPIs and MI

    Dashboards, targets and findings in plain sentences.

  6. Down to the people

    Every figure opens to the records behind it.

  7. Small groups protected

    The smallest group a result may describe is a setting your organisation chooses.

  8. Export anywhere

    Excel, TSV and Power BI.

Employee Relations Cases, types and how long they stay open.
Recruitment: by recruiter The recruiter view: each recruiter's speed, acceptance and whose hires stay.
Survey Desk Load the answers file and where to look first appears: what people say, by group, turned into leads.
Learning and Compliance Mandatory training compliance and training delivered.
9-Boxer A calibration workbook goes out and comes back: the grid draws itself from the ratings in it, and marks the two boxes where WHAT and HOW disagree.
Mapping Desk Unknown columns and wordings are decided here once, one click each, and remembered.

Every picture is the running product on Meridian Bank 2, a synthetic organisation. No real person, client or organisation appears in it. See the full journey, screen by screen, or watch the five-minute demo.

Talk it through

Start with the question your board keeps asking, or the data problem behind it.

Contact me

The other services