Auditable AI systems for UK insurance, health, rehab, and case-based teams.
We help specialist teams turn referrals, evidence, decisions and reporting into clear operational systems — built around existing tools, with human judgement and auditability kept in control.
A measured way to start
Start with one practical handover before committing to a larger build.
The useful discipline is to avoid starting with technology for its own sake. The first engagement should be clear: understand the handover, agree the outcome, test the smallest useful layer, then decide whether further work is justified.
Stage 01
Scoping conversation
We talk through one handover or recurring task, the people involved, the current constraints and whether there is a sensible project to shape.
Stage 02
Diagnostic or process review
Where useful, we map the current state, the outputs people rely on, the audit points and the smallest practical improvement.
Stage 03
Bounded first release
If there is a clear case to build, we agree a focused first release, test it carefully and decide the support model before extending it.
What we do
We reduce the repetitive operational admin — built around how your team works.
AI and custom software that connect the work you already do, cut the manual admin, and keep your people in charge of the decisions that matter.
…and the same skills for the rest of your operation
Case studies
What we’ve built for work, health and specialist case teams.
UK domain grounding
VRA trustee, CII co-author and IRCM member experience informs the way we design for health, rehabilitation and claims work.
Built systems
Cassie and ClaimCompass show the operating layer running in real service pathways, not only as a concept.
Technical delivery
Twenty years of decision-support, data platforms and applied AI behind the systems we build.
Meet the team
The people behind the work.
Monica Garcia
Founder & domain lead
Monica is a strategic consultant to the UK protection sector and founder of Monica Garcia Consulting. Her work across claims, rehabilitation and provider pathways keeps what we build grounded in real operational practice.
Pawel Dobrzynski
CTO & engineering lead
Twenty years designing decision-support systems, data platforms and applied AI for organisations with complex operations. As mgctech’s CTO he leads the technology side — turning domain expertise into practical, auditable systems that teams can actually run.
Common questions
What people usually ask before we start.
Can we start with an initial conversation?
Yes. We usually start with an initial conversation, without charge, to clarify the process, the current constraints and whether there is a practical project worth scoping. It is a working conversation rather than a sales presentation.
What does mgctech actually do?
We design and build practical AI, cloud and operational systems for teams doing complex case-based work. That can mean a new operating layer, decision-support tooling, reporting, automation around existing systems, or a focused diagnostic before anything is built.
How much does a build usually cost?
It depends on scope. Smaller diagnostics and prototypes are priced separately from production systems. For implementation work we prefer fixed, clearly bounded phases so you know what is being delivered before committing further.
Can you work with our existing systems?
Usually, yes. We do not assume you need to replace everything. Many projects start by connecting or improving the tools already in use, then adding the missing process, reporting or decision-support layer around them.
How do you keep AI safe and auditable?
We design AI as support, not as an unchecked decision-maker. Outputs can include source references, review steps, audit trails, role-based access and clear boundaries around what the system may and may not do.
Start
Tell us what you want to improve.
Arrange an initial conversation to clarify the process, the current constraints and the most useful next step. No obligation — simply a practical discussion about what would help.