Francisco Cerdà Puigserver

Background and early career

Initial training in computer engineering. The professional career begins in the late nineties, precisely when digitalization starts to transform the business fabric. That coincidence has shaped the entire trajectory: the work has always taken place at the point of contact between technology and business decisions.

The early projects address problems that today feel familiar but at the time were novel: administrative process automation, integration of heterogeneous systems, the first online stores, infrastructure for small and mid-sized companies. That early experience provides the context needed to understand why certain transformations succeed and others fail.

Building own companies

Over the past fifteen years the activity has concentrated on building and operating digital companies first-hand. This has been a deliberate choice. Advising others without having gone through the process of founding, scaling and sustaining a business yields limited judgment. Operating first-hand forces every decision to be made and every consequence to be lived through.

Two of those projects remain active and constitute the foundation of the current work with clients.

WAzion

WAzion is a SaaS platform for business communication over instant messaging channels. It enables organizations of any size to manage large-scale personalized conversations with their customers, automate commercial and support flows, and coordinate distributed teams. More than two hundred and fifty organizations use it in active production.

Technologically it incorporates a proprietary campaign engine, automatic translation across seven languages, distributed architecture on Node.js and MySQL, and native integration with the major instant messaging channels. The entire infrastructure — development, operations and support — runs with a minimal team, which requires a high degree of internal automation.

My role is founder and technical lead. More information at wazion.com.

NEXO Brain

NEXO Brain is a cognitive architecture designed to operate multiple businesses in parallel. It integrates a long-term semantic memory engine, a specialized autonomous agent system, continuous learning mechanisms from interactions, and persistent synchronization across heterogeneous working environments.

It originates from a concrete need: language models are extraordinarily capable but lack continuity. Every conversation starts without context, every task is solved without memory of prior work, every decision forgets the lessons learned. NEXO Brain is the operational response to that problem.

The system is published entirely as open source. It is likely one of the most ambitious implementations of its kind built by a single operator. Its documentation and source code are available at nexo-brain.com.

Areas of expertise

The advisory work concentrates on six areas, all tied together by a single thread: connecting technology decisions with verifiable business outcomes.

Technology strategy: supporting executive teams on decisions about technology adoption, architecture structure and investment prioritization.

Process automation: identification, design and implementation of automation across critical operational processes. Pragmatic and tool-agnostic.

Artificial intelligence integration: from use-case identification to production deployment with observability. Autonomous agents, semantic retrieval, automated evaluation.

Data infrastructure: design of data architectures that enable informed decisions without accumulating unnecessary complexity.

Advanced e-commerce: ecommerce operations with optimized advertising, automated support and unit-level margin measurement.

Technical leadership: supporting founders and CEOs on structural technology and product decisions.

Sectors and organization types

The diversity of contexts I have worked in over the years is probably one of the most valuable assets I bring to each client. I have collaborated with organizations of deliberately varied profiles: tourism companies with high seasonal variability, industrial multinationals with complex decision structures, e-commerce operators running dropshipping models, small niche family businesses, service corporations with thousands of employees, and early-stage startups searching for product-market fit.

This diversity is not accidental. Working across such different sectors reveals what is universal and what is circumstantial: good processes resemble each other regardless of what the company sells, and mistakes tend to repeat with the same structure. An automation failure at a hotel chain is rarely different from an equivalent failure at a logistics operator or at a professional services firm.

Confidentiality is non-negotiable. Out of respect for the agreements signed, no named references to current or past clients appear on this page or in any other public material. Concrete details that allow a client to assess the fit can be discussed privately during the first conversation.

Working methodology

Every engagement shares the same structure, refined through years of practice: in-depth analysis of the business model, prioritized diagnostic of initiatives, execution with direct supervision, and continuous measurement through verifiable business indicators. This structure is what allows consistent delivery regardless of the client's sector.

Commitments are deliberately concentrated in a limited number of active projects at any point in the year. This restriction is a necessary condition for maintaining the quality of attention this kind of work requires.

Format and scope

All work is conducted in a fully remote format. Communication is held in Spanish or English, at the client's preference. There is no geographic restriction: engagements are open to organizations anywhere in the world.

Confidentiality is the foundation of every working relationship. Non-disclosure agreements are signed as a prerequisite to any exchange of sensitive information and, consequently, no named references to current or past clients appear on this page.

Technical stack and tools

Advisory work requires real technical capability. The following is a list of the technologies and tools I work with directly, regularly and in production. It is not an aspirational list: all of them are in active use in own or client projects right now.

Programming languages: Python, JavaScript and TypeScript, PHP, SQL, Bash, HTML/CSS.

Runtimes and frameworks: Node.js, Express, FastAPI, Flask, Astro, Next.js.

Databases: PostgreSQL (including pgvector), MySQL, SQLite, Redis. Schema design, query optimization, replication and partitioning.

Infrastructure and systems: Linux (Debian, Ubuntu), Apache, Nginx, Docker, SSH, systemd, launchd, cron. Deployment on own servers, cPanel, Google Cloud Platform and AWS.

Artificial intelligence and language models: Anthropic (Claude), OpenAI (GPT), Google (Gemini) and open source models (Llama, Mistral). Semantic retrieval (RAG) with vector stores, automated evaluation systems, tool-using autonomous agents, embedding pipelines and fine-tuning when applicable.

Automation and integration: Make.com, n8n, Zapier, direct APIs, webhooks, scheduled cron scripts. Integration with external services via OAuth, tokens and custom authentication.

Marketing, analytics and e-commerce: Shopify, WooCommerce, Google Ads, Meta Ads, LinkedIn Ads, TikTok Ads, Microsoft Advertising, Google Analytics 4, Google Search Console, Google Tag Manager, Meta Pixel, Conversions API.

Messaging and communication: WhatsApp Business API, Twilio, integration with instant messaging channels, transactional email (SMTP, SES, Mailgun).

Version control and collaboration: Git, GitHub, GitLab. Branch management, continuous integration, automated deployment.

Contact

To begin a no-commitment conversation: info@systeam.es or +34 648 903 026. Response within twenty-four hours.