AI Solutions Architect

Eduard Reñé Claramunt

Helping teams ship production-ready AI.

Architecture and advisory for enterprise AI, agents, RAG, and LLMOps. Founder of Global Insights.

Discuss an AI initiative View selected work Global Insights →

The person behind the architecture

About

I help CTOs, engineering leaders, and founders make sound decisions about AI systems—then design architectures that can actually run in production.

Journey

I trained as a computer engineer in Catalonia, contributed to Drupal 8 core in Zürich, built mobile products in Poland, and spent years in Stockholm shipping platforms and AI capabilities at Digitalist. That path led to founding Global Insights in Barcelona: an AI engineering consultancy focused on production systems and team enablement.

Engineering philosophy

Architecture before models. Evaluation before demos. Privacy as a default, not a retrofit. Capability should remain with the client's team after the engagement—not locked in a black box only I can operate.

AI expertise

I work across LLM applications, agents, enterprise RAG, and classical ML where it still wins. The hard problems are usually systems problems: data boundaries, integration, evaluation, failure modes, and operability.

Architecture mindset

I design for integration into existing stacks, clear ownership, graceful failure, and long-term maintainability. Fancy prototypes that cannot be owned by your engineers are not a success.

Consulting approach

Start with discovery and constraints. Separate judgment (advisory) from delivery (often via Global Insights). Prefer proof-of-concepts that answer a real risk question before committing to a full build.

Leadership

As a Scrum Master and senior engineer I learned that delivery leadership is about standards, clarity, and enabling others—skills I now apply to AI engineering enablement.

Based in Barcelona

I work in Catalan, Spanish, and English, and bring an EU privacy-aware default to every engagement.

Capabilities, not keyword clouds

Expertise

How I apply AI engineering, architecture, and enablement in real engagements.

Enterprise RAG

Retrieval systems with clear data boundaries, evaluation, and operability.

AI Agents

Agent patterns that are reliable enough for real workflows—not fragile demos.

Architecture & advisory

System design, integration, and trade-offs before models become the bottleneck.

MLOps / LLMOps

Evaluation, monitoring, and release practices so AI stays trustworthy in production.

AI Engineering

Building LLM applications and intelligent features with engineering discipline.

  • LLM applications
  • AI agents
  • Retrieval-augmented generation
  • Evaluation design
  • Classical ML when it fits

In practice: I connect model capabilities to product workflows, with evaluation and failure modes defined early.

Software Architecture

Systems that integrate, scale, and can be owned by your team.

  • Integration into existing stacks
  • Modular boundaries
  • API design
  • Architecture trade-offs
  • Clear ownership models

In practice: I optimize for clear boundaries, operability, and honest trade-off conversations with technical leaders.

Enterprise RAG

Retrieval systems that respect data boundaries and stay measurable.

  • Retrieval design
  • Data boundaries & access
  • Grounding & citations
  • RAG evaluation

In practice: I treat RAG as an information architecture problem first—then a model problem.

AI Agents

Agents that are useful under real constraints.

  • Agent patterns
  • Tool use design
  • Reliability & guardrails
  • Governance

In practice: I prefer narrow, evaluable agents with explicit tools and governance over open-ended autonomy theater.

MLOps / LLMOps

Operating AI after the demo works.

  • Monitoring
  • Evaluation harnesses
  • Prompt/model release practices
  • Cost & latency awareness

In practice: I help teams instrument quality, cost, and latency—and release prompts/models with intent.

Developer Experience & Enablement

Making engineering organizations AI-native without sacrificing quality.

  • AI-assisted workflows
  • Engineering standards
  • Team coaching
  • Adoption programs

In practice: Workflows, standards, and coaching so AI adoption compounds inside the team.

Backend, Cloud & Operability

Enough depth to own deployment and integration conversations.

  • Node.js & Python systems
  • Containerized delivery
  • Cloud & self-hosted options
  • Operability concerns

In practice: Node.js and Python delivery, containers, and cloud or self-hosted options matched to security needs.

Projects & products

Selected work

I am building my consulting practice. Here is honest selected work: products offered through Global Insights, a work-in-progress, and AI delivery from my previous role—not a long agency portfolio.

SaaS · Global Insights

Talaia — trademark watch for OEPM

An AI-assisted watch service that monitors Spanish OEPM brand filings for conflicts against a client watchlist—offered as SaaS by Global Insights.

Read case study

Context

Brand owners need early warning when new trademark filings at the Spanish Patent and Trademark Office (OEPM) may conflict with marks they already protect. Manual monitoring does not scale.

Approach

Build a watch pipeline over OEPM submissions, match against a curated watchlist, and surface likely conflicts so teams can act early. Deliver it as a Global Insights SaaS product rather than a one-off script.

Role

Product and AI design for the watch logic, conflict signalling, and how the service is offered commercially through Global Insights.

Outcome

Talaia as a live SaaS offering: continuous monitoring for trademark conflict risk on OEPM filings.

Lessons

Useful AI in legal-adjacent workflows is about reliable monitoring and clear alerts—not flashy demos.

Visit Global Insights

SaaS · Global Insights

Map server — geospatial SaaS

A map server product sold as SaaS by Global Insights for teams that need dependable map infrastructure without building it from scratch.

Read case study

Context

Many products need maps, tiles, and geospatial serving, but owning that stack end-to-end is expensive and distracts from the core product.

Approach

Package a maintainable map-server capability as a Global Insights SaaS offering: operable, documented, and ready for customer workloads.

Role

Architecture and product definition for the service model, operability expectations, and how it fits the Global Insights catalog.

Outcome

A SaaS map-server offering clients can adopt instead of standing up fragile one-off infrastructure.

Lessons

Consultancy products should solve recurring infrastructure pain with clear ownership—not custom snowflakes every time.

Visit Global Insights

In progress

Pharos Explorer — curated human guides

Work-in-progress application for discovering curated, human-authored guides. AI helps creators polish content—it does not replace the guide.

Read case study

Context

People still want trustworthy, human-curated guides. Generative AI can help editors improve clarity, but the value is the human expertise behind the guide.

Approach

Design an application around curated guides and creator workflows. Use AI as a polishing assistant for authors, keeping humans in charge of substance and curation.

Role

Product direction and AI-assist design for the creator experience while the product is still in progress.

Outcome

A clear product direction in progress: human guides first, AI as editorial assistance—not an auto-generated content mill.

Lessons

Positioning AI as helper to experts is often more credible than claiming AI replaces expertise.

Past role

AI engineering at Digitalist

Applied AI inside a product organization: LLMs, agents, NLP tooling, and AI-assisted analytics—built while shipping real software platforms.

Read case study

Context

Digitalist needed AI capabilities that marketers and operators could use in daily workflows—not isolated experiments.

Approach

Introduce LLM and NLP tooling, design agents carefully, and expose insights through reports and dashboards. Pair that with years of platform engineering (multisite, APIs, headless, mobile).

Role

AI engineer and previously software developer / Scrum Master—delivery leadership plus hands-on AI implementation.

Outcome

AI features embedded in product and analysis workflows, grounded in a strong software platform background.

Lessons

Production AI inherits the quality of the underlying engineering culture and architecture.

Live

Founding Global Insights

Building the AI engineering consultancy that offers production systems, enablement—and products like Talaia and the map server.

Read case study

Context

Organizations need production AI and clear productized services, not only advisory slides.

Approach

Found Global Insights around delivery and enablement, then productize repeatable capabilities (watch services, map infrastructure) as SaaS where it helps clients.

Role

Founder—positioning, service design, product direction, and client advisory.

Outcome

A consultancy home for team delivery, complementary to personal advisory on this site.

Lessons

Personal brand and company brand should reinforce each other: judgment here, delivery and SaaS there.

Visit Global Insights

Public

Open source craftsmanship

Drupal 8 core and module contributions—public engineering discipline that still informs how I review systems.

Read case study

Context

Large CMS ecosystems demand careful APIs, compatibility, and community process.

Approach

Contribute upstream and learn from review culture on a global project.

Role

Drupal 8 developer and contributor.

Outcome

A public contribution history and habits of rigor.

Lessons

Open source proves craft; it is supporting evidence for an architect brand, not the whole story.

Visit Global Insights

Selected chapters—not a full CV

Experience

Highlights that demonstrate architecture, AI delivery, leadership, and international practice.

  1. 2018 – 2024

    Stockholm, Sweden

    Software platforms → AI engineering

    Digitalist Open Tech AB

    Over six years I helped design and deliver complex web platforms—then led AI capabilities into product workflows: LLMs, NLP, agents, and operational tooling.

    • Architected multilanguage, multisite platforms with a single deployment model.
    • Built headless and API-driven systems (Node.js, React, Symfony) and mobile clients.
    • Introduced LLM-based features, agents, and AI-assisted analytics into real product contexts.
    • Served as Scrum Master—delivery leadership, facilitation, and quality standards.

    Demonstrates: Architecture AI expertise Leadership Product thinking

  2. 2015 – 2016

    Zürich, Switzerland

    Drupal 8 engineer & core contributor

    MD Systems Miro Dietiker

    In Zürich I worked on Drupal 8 and contributed to core and modules—deep craftsmanship in large open-source systems.

    • Contributed to Drupal 8 core and the module ecosystem.
    • Practiced rigorous engineering inside a demanding international team.

    Demonstrates: Architecture Open source Software engineering

  3. 2017

    Zaragoza, Spain

    Architecture & delivery across stacks

    ENDPHASYS Technologies

    Short, intensive delivery spanning Drupal commerce, Symfony services, and data acquisition pipelines.

    • Designed and built a Drupal 8 e-commerce architecture.
    • Delivered Symfony services and scraping pipelines with operational constraints in mind.

    Demonstrates: Architecture Software engineering

  4. 2014

    Łódź, Poland

    Mobile product engineering

    Brainsupport

    Designed and developed iOS applications while owning delivery in an international setting.

    • Built iOS apps with Swift and Objective-C.
    • Combined hands-on engineering with project ownership.

    Demonstrates: Product thinking Software engineering

A detailed CV is available on request for recruiting conversations.

Senior judgment for complex AI initiatives

Advisory

Personal offerings for architecture clarity, fractional capacity, and technical advisory. Team build-and-operate work is delivered through Global Insights.

Work with me

Architecture reviews, fractional AI architect roles, technical advisory, and diligence.

Work with Global Insights

When you need a team to design, build, and operate production AI systems—or enable your engineers.

Flagship engagements

Architecture review

A focused review of your AI system or initiative: integration, risk, evaluation, and a concrete target architecture.

Ideal for
CTOs and tech leads before or after a pilot.
Outcome
Written findings, recommended architecture, and prioritized next steps.
Typical shape
Several concentrated days
Discuss this engagement

Fractional AI Architect

Embedded senior architecture capacity without a full-time hire—decision support, design, and delivery oversight.

Ideal for
Startups and product teams shipping AI into the core product.
Outcome
Ongoing architectural ownership and clearer delivery decisions.
Typical shape
Part-time retainer
Discuss this engagement

Technical advisory retainer

A sounding board for CTOs and engineering managers on AI strategy, platform choices, and risk.

Ideal for
Leaders who need recurring judgment, not a full project team.
Outcome
Faster, safer decisions with documented recommendations when needed.
Typical shape
Monthly retainer
Discuss this engagement

Technical due diligence

Independent assessment of AI vendors, stacks, team readiness, or acquisition targets.

Ideal for
Buyers, investors, and corporate development teams.
Outcome
Diligence memo with risks, gaps, and go/no-go considerations.
Typical shape
Scoped days
Discuss this engagement

AI engineering enablement sprint

Raise how your engineering organization adopts AI: workflows, evaluation, tooling, and guardrails.

Ideal for
Engineering managers building AI-native practices.
Outcome
Practical operating model your teams can own.
Typical shape
Multi-week sprint
Discuss this engagement

Also available

  • AI strategy workshops
  • AI solution design
  • Enterprise AI advisory
  • AI agent design & critique
  • Enterprise RAG architecture
  • MLOps / LLMOps assessment

Let's talk

Contact

Tell me what you are trying to achieve. I reply within a few business days.

Architecture reviews, retainers, diligence, enablement.

Enable JavaScript to view email

Barcelona, Spain

Typical response: within 3 business days.

Send email