Ian Calvert / Independent technical consultant

Technical judgement.
Working code.

I help teams work through difficult data, AI and software problems — and build what’s needed to test the answers.

Discuss a project

From the first difficult question
to a practical next step.

Experience across
software, data & research

BBCSoftware engineering & testing
Digital ScienceHead of Data Science
Independent workApplied AI & data projects

Make the difficult part
something you can test.

Some questions need more than a specification. They need someone who can get into the data, build an approach and find out where it holds up.

Build

Prototypes with a purpose

Test a product idea or technical approach with real data and users. Build enough to answer the question before committing a larger team.

Connect

Data that supports decisions

Build analytics, ML and AI around the decisions people need to make. Establish whether the underlying data is reliable, test the approach and judge its value in day-to-day use.

Investigate

A clear view of the problem

Get inside an existing system. Trace unexpected behaviour, evaluate models or rules, and establish which changes are worth making.

Design

A sound technical direction

Work through architecture, product and engineering decisions with your team, grounded in what the code and data actually show.

About Ian

My background spans software engineering and data science, including software delivery for the BBC’s Olympics coverage and leading data science at Digital Science. I’ve worked on products used by millions and tools built for a single person’s work.

That experience informs the questions I bring to a project: whether the data supports the proposed use, how to evaluate the results, and whether a technical improvement will make a useful difference to the business.

I work directly with clients on those questions and the software needed to answer them. My work covers investigation, design and implementation, with continued involvement as projects develop.

Start with the question.
Leave with a way forward.

01 /

Understand the decision

What do you need to know? We establish the question, the constraints and what would count as a useful answer.

02 /

Build, test, investigate

I work with your data, code and team. We test the uncertain parts, measure performance and make the trade-offs visible.

03 /

Make the next step clear

You get working code, the evidence behind it and recommendations your team can build on.

What are you
trying to work out?

Tell me a little about the problem and where you’re stuck.
It doesn’t need to be a finished brief.