Principal solutions architecture · Technical R&D leadership
Make the difficult
make sense.
Then make it work.
I turn unclear business problems into systems that people can understand, build and rely on.
- 20+ years
- Technical problem-solving and systems across live production, automation and software.
- 7+ years
- Web development and client delivery, from hands-on builds to technical leadership.
AI-native, with human judgement at the centre. AI is built into how I investigate, design, build and verify. I bring the systems understanding, set the direction and stay accountable for the result.
Philip Carlsson · Swedish · Based in Vietnam · Remote, international work
Systems problem-solver.
Hands-on maker.
01 / Demonstrated value
The work is the strongest introduction.
A consistent pattern across live production, independent client delivery and technical leadership: understand the real constraint, make it solvable and stay accountable for what follows.
I build capability around the problem, not the job title.
Live systems work and independent client delivery established this way of working long before I led a technical function. At Fortis, the same pattern grew into R&D leadership, platform direction and developer support.
Systems diagnosis · Client delivery · Technical leadershipExplore my experience Direct engineering / 2026When the policy would not survive a handoff, I implemented the core.
A C#/.NET work-time engine brings activity intervals, time zones, mixed project thresholds, lunch and gap recovery into an explainable day-level model.
Direct policy and implementation work · Later structural work by the teamExplore the project Architecture / three generationsThe product outgrew its provider. The architecture had to change.
Led a core operational platform from spreadsheet-proven rules through React and Blazor generations toward a canonical model serving multiple operating systems.
Domain and architectural direction · Developer-led application implementationExplore the project Measured data engineeringA pipeline shaped around completion, not batch barriers.
Six dependent PowerShell stages with bounded pressure and live telemetry. Database evidence records 502,253 distinct IDs completed on 23 June 2024, with zero duplicate IDs.
Measured production run · June 2024Explore the project Production integrationNo suitable integration? Understand the actual mechanism.
Reconstructed browser-compatible authentication and built durable recruitment synchronization across systems with different definitions of a candidate. About 60 applications per day in the described workflow.
Direct engineering · Recruitment workflow automationExplore the project Commercial ownershipA technical solution is only part of a deliverable engagement.
Turned an international website opportunity into a signed engagement with discovery, scope, migration, specialist responsibilities, acceptance and delivery risk made explicit.
Discovery · Signed scope · Delivery planningExplore the project02 / Why this combination matters
One accountable mind across the boundaries.
The advantage is continuity. The business meaning survives the move into data, code, infrastructure, team decisions and live operation.
Systems architecture
Find the governing model.
When interfaces disagree or several systems own a fragment of the truth, I work out what the business actually needs to mean.
Architecture and range →Business-rule design
Make the judgement explainable.
Fair attribution, historical reversals and qualified time need more than a convenient threshold. Rules must survive real counterexamples.
Rules and temporal reasoning →Leadership
Give people a model they can own.
I turn ambiguous goals into direction, evidence and acceptance criteria. I delegate implementation without losing responsibility for the intent.
Leadership in practice →Difficult engineering
Go beneath the abstraction.
Protocols, concurrency and failure paths are often where the real constraint lives. I investigate and directly solve selected high-complexity boundaries.
Engineering depth →AI-native delivery
Turn experience into execution leverage.
I turn operational knowledge into clear models, precise AI instructions and working software. The advantage is knowing what to build, what to challenge and how to tell when it is right.
How I work with AI →
Think it through.
Then build it properly.
From the first working model to the details that make it dependable. The tools can change. Responsibility for the result stays with me.
03 / How I think
Start with reality.
Not the favourite tool.
I am most useful before the problem has become a neatly specified ticket. What looks like a missing feature may be a false data model, a policy conflict or an unowned operational boundary.
The method, in depth →- Reconstruct the problemTrace the workflow and find where the facts disagree.
- Expose the ruleSeparate source behaviour, policy and assumptions.
- Prove the hard partUse examples, executable models and counterexamples.
- Lead it into operationMatch engineering and verification to consequence.
04 / Range with a foundation
Hardware. Software.
One working system.
In concert and broadcasting work, software and scripting were already part of the system: automating tasks, connecting equipment and configuring hardware. Live audio and show systems also demanded that I trace complete signal paths under immediate pressure. Independent web delivery taught me to scope, hire and lead. Operational tooling taught me where elegant software meets messy working reality.
That foundation now reaches across .NET and Blazor, PowerShell, APIs, data engineering, Google Apps Script, infrastructure and applied AI. The tool changes. The habit of understanding the whole system does not.
The maker and live-systems foundation →“Knowledge is what you know, Intelligence is how you fill in the blanks.”
05 / The right mandate
Hire me for the problem. Not just the stack.
Strongest fit: Principal Solutions Architect, Technical R&D Lead or Technical Client Solutions Lead. A hands-on Head of Technology / R&D mandate also fits a small or mid-sized organisation.
Where I create the most value
Ambiguous cross-system problems. Internal platforms that need a real operating model. Technical functions that need direction and ownership. Client opportunities that need to become well-scoped, dependable delivery.
What the mandate needs
Room to investigate, authority to challenge a false premise and a team or specialist capacity to execute. I create the most value when I can connect technical decisions to the wider business outcome.
Roles and responsibilities in depth →Start with the actual problem
What needs to become possible?
Tell me the role, the situation and where it is getting stuck.