Software and AI engineering
Systems are designed as complete products, not isolated model calls. Architecture, data, interfaces, evaluation and deployment are considered together.
About Norrelium
Norrelium is led by Derril Filemon, an AI systems engineer and product-minded builder with experience across software engineering, business management, design, leadership and real operational environments.
01 / DETAIL
Powerful AI models are widely available. Dependable business systems are not. Norrelium was founded to close the gap between technical possibility and operational reality.
02 / DETAIL
Derril Filemon is Founder and AI Systems Engineer at Norrelium. Before establishing the company, he worked independently across AI application development, agentic workflows and full-stack software engineering. That work established the technical foundation, delivery process and project portfolio that became Norrelium.

03 / DETAIL
A multidisciplinary background helps connect commercial priorities, operating constraints and product decisions to the engineering work.
Systems are designed as complete products, not isolated model calls. Architecture, data, interfaces, evaluation and deployment are considered together.
Technical decisions are tied to commercial priorities, ownership and measurable outcomes. Leading and coaching teams strengthened Derril’s ability to connect performance data, communication and practical improvement.
Experience in high-volume operational environments strengthened Derril’s focus on accuracy, traceability, exceptions and reliable execution.
Design and product experience help turn complex AI behaviour into interfaces and workflows people can understand and use.
Experience handling complex, cross-functional issues supports a root-cause approach rather than treating only the visible symptom.
04 / DETAIL
Clients work directly with Derril from the first business discussion through technical delivery. This preserves context, reduces handoffs, supports faster decisions and keeps accountability with the person shaping and building the system—without junior-team substitution.
05 / DETAIL
Credibility is presented by discipline rather than as a wall of certificates. The external portfolio provides the inspectable detail behind this foundation.
Inspect working interfaces, technical decisions and clearly labelled independent projects.
Review the broader education, experience and delivery background supporting Norrelium.
View the maintained external record rather than duplicating a full certification catalogue here.
06 / DETAIL
Norrelium’s working method is deliberately practical and evidence-led.
Understand the user, workflow and desired result before selecting technology.
Make architecture, trade-offs and failure behaviour understandable.
Place review and approval where ambiguity or consequence requires a person.
Use representative scenarios, acceptance criteria and failure cases.
Document decisions and make operating responsibility explicit.
Adopt new methods when they improve the outcome—not because they are new.
07 / DETAIL
Based in Sweden, with professional and personal experience across the Nordics and Benelux. Derril works across multiple European languages, including English, Dutch, Norwegian and Swedish.
08 / DETAIL
Norrelium’s independent engineering work makes architecture, interfaces, control mechanisms and limitations inspectable.
A source-grounded research interface demonstrating retrieval, ranking, streaming answers and visible citations.
A controlled agent workflow demonstrating visible plans, confirmation gates and observable execution.
Next step
Bring a clear bottleneck, an early AI idea or an existing prototype. Derril will help establish the business context, technical fit and smallest credible next step.