21.11.26

What to Look for When Hiring a Technical Project Manager for a Regulated Data or AI Programme

Why a Generic PM Won't Cut It on a Regulated Data or AI Programme

Hiring a Technical Project Manager for a regulated data or AI programme is not the same brief as hiring a delivery manager for a generic IT rollout. The role sits between architects, data scientists, and compliance stakeholders, and the wrong hire shows up as missed dependencies, a program lead firefighting detail that should never have reached them, and a delivery team that stops trusting its own plan. We've run enough of these conversations with candidates and hiring managers to know exactly where the generalist profile breaks down — and exactly what good looks like instead. Here's how to run the hiring process so you land the right one.

Why a Generic PM Won't Cut It on a Regulated Data or AI Programme — dcoded. Manchester tech recruitment

Step 1: Define the Role as Technical-First, Not Change-First

Before you write a single interview question, get clear internally on what this person actually does day to day. A regulated data or AI programme typically splits into two layers: a program lead who owns strategic stakeholder engagement, and a Technical Project Manager who owns day-to-day coordination of architects, analysts, and engineers, making sure delivery is on track and built to a defined standard and timeline.

Step 1: Define the Role as Technical-First, Not Change-First — dcoded. Manchester tech recruitment

That second layer is the one hiring managers most often get wrong, because it reads on a job spec like a standard PMO role. It isn't. The remit includes making sure a program lead's strategic direction is actually being delivered against day-to-day by the data teams underneath it. If your candidate can't speak fluently about pipelines, cloud environments, and how a data team actually works, they can coordinate meetings but they can't catch the things that go wrong between them.

Note: if your job description reads like it could apply equally to a construction PM or an office-move coordinator, rewrite it before you start sourcing.

Step 2: Test for Genuine Cloud Fluency, Not Just Cloud Exposure

A credible Technical Project Manager for this type of programme needs to have actually worked inside a cloud environment, not simply managed a project where cloud was mentioned in the scope. Look for candidates who have been hands-on with platforms like Azure, including working with Azure DevOps, Azure pipelines, and CI/CD, and who have worked closely with architects early in a project to ensure the platform and environments are in place before developers start building.

Step 2: Test for Genuine Cloud Fluency, Not Just Cloud Exposure — dcoded. Manchester tech recruitment

The strongest signal is a candidate who can walk you through a specific build, not a program overview. One benchmark example from our own candidate conversations: a Technical Project Manager who worked with an Azure Microsoft Gold Partner on a big data project, prototyping a legal case database in Azure using Azure Databricks and Gremlin queries, loading roughly a million data points from the original Enron case load and linking it via nodes and edges in an early AI-style database build. That's the level of in-the-weeds technical detail that separates someone who ran a cloud project from someone who understood it.

  • Ask for a specific cloud project, not a cloud-adjacent one.

  • Ask what role they played with the architects before development started, not just during delivery.

  • Ask whether their cloud experience is Azure-specific or whether it transfers across providers — strong candidates will tell you cloud environments are broadly similar once you're working at that level.

This also matters if your own programme sits on GCP or another provider rather than Azure. Candidates coming from complex, regulated cloud projects — even on a different platform — bring a transferable discipline around environments, pipelines, and architecture sequencing that a generalist delivery manager simply won't have built.

Step 3: Probe Their Working Relationship With Data Science Teams

Cloud literacy on its own isn't enough. The next filter is whether the candidate has actually worked alongside a data science function, not just managed a project that happened to contain one. In government data environments, for example, the strongest Technical Project Managers have worked closely with data scientists delivering data reporting for clients, and in some cases have sat inside the data science team itself rather than managing it from the outside.

Step 3: Probe Their Working Relationship With Data Science Teams — dcoded. Manchester tech recruitment

This distinction matters because regulated data and AI programmes are moving steadily from reporting into AI-driven delivery, and the structured release processes that govern that shift — in government, for instance, following the GDS standard through alpha, beta, and live — require a Technical Project Manager who understands why that structure exists, not just how to track it on a plan.

When you're assessing this in interview, ask the candidate to describe:

  • A specific data team they coordinated — who was on it (engineers, architects, analysts), and how they kept delivery on track across that mix.

  • Whether they were managing data scientists from a PM seat, or embedded enough in the team to understand the technical trade-offs the scientists were making.

  • How their coordination changed when a project moved from a reporting-focused phase into something closer to AI delivery.

A candidate who can only describe "running the meetings" and "owning the RAID log" is describing PMO administration, not technical delivery management. The stronger profile owns the project plans and chairs the meetings, but can also speak the language of the data professionals they're coordinating through delivery.

Step 4: Weight Regulated-Sector Experience Correctly — FCA and Beyond

For financial services and public sector hiring managers specifically, don't treat "regulated experience" as a single checkbox. Dig into what kind of regulated environment the candidate has actually operated in, because the disciplines differ. A candidate may have direct experience working with the FCA without financial services being their deepest area of expertise overall — which is a perfectly legitimate profile, provided they can also point to other highly structured, highly organised environments where they've operated against defined implementation protocols, such as government delivery.

What you're really testing for is whether the candidate understands why regulated environments move more slowly and require more rigour — not just that they've sat in a room where regulators were mentioned. Someone with FCA exposure plus government delivery experience, for example, has been tested on structured, protocol-driven delivery from two different angles, which is a stronger signal than FCA experience alone with no other regulated context to compare it against.

For a public-sector-facing hire, ask specifically whether the candidate has delivered against a named standard or framework (the GDS standard and its alpha/beta/live process is a clear, checkable example) rather than a vague reference to "working with government clients".

Step 5: Use a Structured Interview to Separate Signal From Spec-Matching

By this stage you should have a shortlist of candidates whose CVs look aligned. The risk at interview is letting a confident narrative substitute for a concrete one. Structure the conversation to force specificity:

  1. Open with the project, not the title. Ask the candidate to talk you through a specific complex data or cloud project they delivered, including what was built, on which platform, and what their actual remit was day-to-day.

  2. Push past the headline tech stack. If they say "Azure," ask what specifically — pipelines, CI/CD, Databricks, a named build. The difference between a candidate who says "I worked in Azure" and one who can describe prototyping a Databricks-based build against a dataset of a known scale is the difference you're hiring to find.

  3. Ask who they worked with, not just what they delivered. A strong answer names the mix: architects up front, developers through build, data scientists throughout, and a program lead above them handling strategic stakeholder engagement while they owned day-to-day technical coordination.

  4. Test regulated-environment reasoning. Ask what changed about how they ran delivery because the environment was regulated — structured protocols, a formal release standard, heavier sign-off — rather than accepting "it was regulated" as a complete answer.

  5. Confirm hands-on delivery ownership. Strong candidates will describe owning the project plans, chairing project meetings, and owning the RAID log themselves, not delegating that coordination while taking credit for the outcome.

Tips & Best Practices

  • Hire for the seam, not just the skillset. The Technical Project Manager's real value is in the gap between a program lead's strategy and a data team's execution — test candidates on how they've closed that gap before, not just on tools they've used.

  • Favour depth over breadth of regulated exposure. One deeply understood regulated environment, explained with real protocol detail, is a stronger hire signal than a CV that lists multiple regulators in passing.

  • Don't assume Azure-only experience is a blocker for a non-Azure programme. Strong technical PMs will tell you directly that core cloud discipline — environments, pipelines, architecture sequencing — transfers across providers.

  • Listen for whether they were "in" the data team or managing it from outside. The best signal of genuine fluency with data science teams is having actually sat inside one, not just overseen it from a project-plan distance.

Troubleshooting: Common Hiring Mistakes

  • Mistaking change management certification for technical capability. Certifications like Prosci are valuable for a change manager role, but they test a different discipline entirely from hands-on technical coordination of data and cloud delivery — don't let a change-management background substitute for proven technical delivery experience on this kind of hire.

  • Accepting "I managed the stakeholders" as a complete answer. It's necessary, but on its own it describes the program lead layer of the role, not the Technical Project Manager layer you're actually hiring for.

  • Over-indexing on sector match and under-indexing on structure match. A candidate from a differently regulated sector who has operated under an equally rigorous structured protocol can be a stronger hire than one with light-touch exposure to your exact sector.

Summary

Hiring a Technical Project Manager for a regulated data or AI programme means hiring someone who can sit credibly between architects, data scientists, and a program lead's strategic agenda — someone with real cloud build experience, real time spent working alongside data teams, and a genuine understanding of why regulated environments demand the structure they do. Test for specifics at every stage: the named platform, the actual build, the data team they coordinated, and the protocol they delivered against. Get this hire right, and your delivery team gains a layer of technical assurance that a generalist delivery manager was never built to provide.

What to Look for When Hiring a Technical Project Manager for a Regulated Data or AI Programme

Latest News