What we build and how we help
Happyberg Labs is a tech boutique that helps smart teams ship better products. We combine 20+ years of engineering experience with deep expertise in AI, digital health, and product development. Based in Galicia, Spain, serving teams worldwide.
AI and Machine Learning Development
We build production AI systems that solve real business problems, not proof-of-concept demos that never ship. Our AI work spans agentic workflows, computer vision, natural language processing, and custom machine learning models.
We specialize in building agentic AI systems using LangChain and LangGraph, where autonomous agents handle complex multi-step workflows. We also build computer vision pipelines with MediaPipe and PyTorch for tasks like medical imaging analysis, lab automation, and quality inspection.
Our approach to AI is pragmatic: we start with the simplest solution that could work, validate with real data, and iterate. According to Gartner, by 2028, 33% of enterprise software will include agentic AI. We help teams get there with production-ready systems, not hype.
- Agentic AI - Multi-agent workflows with LangChain, LangGraph, and custom orchestration
- Computer Vision - Medical imaging, lab automation, real-time video analysis with MediaPipe and PyTorch
- NLP and Embeddings - Semantic search, document understanding, RAG systems with vector databases
- Custom ML Models - Supervised and unsupervised learning, time series analysis, anomaly detection
Fractional CTO and Technical Leadership
A fractional CTO provides senior technical leadership on a part-time basis. This model gives startups and SMEs access to experienced technical direction without the full-time cost, which typically saves 60-70% compared to hiring a full-time CTO.
In practice, fractional CTO engagement means 2-3 days per week of hands-on involvement: setting technical strategy, mentoring your engineering team, making architecture decisions, reviewing code, and managing technical debt. We also help with hiring, vendor evaluation, and investor-facing technical due diligence.
This service is ideal for companies with 3-30 engineers who need senior leadership but are not ready for a full-time executive hire. We have provided fractional CTO services across digital health, SaaS, fintech, and research-driven organizations.
| Aspect | Full-time CTO | Fractional CTO |
|---|---|---|
| Cost | $200-400K+/year | 30-40% of that |
| Availability | Full-time, single company | 2-3 days/week, focused |
| Experience breadth | Varies | Cross-industry, multi-stack |
| Commitment | Long-term hire | Flexible engagement |
Digital Health and Medical Systems
Digital health is a core specialization at Happyberg Labs. Our founder, David Gil Pérez, has co-founded two digital health companies and has over a decade of experience building medical software that meets clinical requirements and regulatory standards.
We build telemedicine platforms, medical imaging systems that handle DICOM data, clinical workflow automation, and EHR integration using HL7 and FHIR interoperability standards. Our telemedicine work has connected healthcare facilities across multiple continents, including a project linking Antarctica with the International Space Station for remote clinical consultations.
Healthcare software has unique constraints: regulatory compliance, data privacy (GDPR, HIPAA), interoperability standards, and zero tolerance for errors. We understand these constraints from years of building systems that run in hospitals and research labs.
- Medical Imaging - DICOM viewers, image analysis pipelines, PACS integration
- Telemedicine - Video consultation platforms, remote diagnostics, WebRTC-based systems
- Interoperability - HL7 v2/v3, FHIR R4, EHR integration, clinical data exchange
- Clinical Workflows - Lab automation, specimen tracking, result reporting
Product Development and Strategy
We help teams move from idea to shipped product using proven methodologies. Our approach combines Shape Up for scoping work in 6-week cycles, Jobs to Be Done (JTBD) for understanding what users actually need, OKRs for aligning work with measurable outcomes, and YAGNI to avoid building what is not needed yet.
Product development at Happyberg Labs means choosing the right technology for the problem, building the minimum needed to validate assumptions, and shipping consistently. We ship something every Friday. This cadence forces decisions and prevents the drift that kills products.
We work across the full product lifecycle: from pitch deck to production deployment, from first user to scale. We have built SaaS platforms, internal tools, research applications, and data pipelines for teams ranging from 2-person startups to 50-person engineering organizations.
Infrastructure and DevOps
We design, build, and manage infrastructure across cloud and on-premise environments. Our infrastructure work spans PostgreSQL and MySQL database administration, Docker containerization, Linux server management, network architecture, and both AWS and Azure cloud deployments.
We also maintain bare metal servers for workloads where cloud costs are prohibitive or latency requirements demand physical proximity. Our current infrastructure manages Docker Swarm clusters running 50+ services, event-driven architectures with RabbitMQ, monitoring stacks with Grafana and VictoriaMetrics, and VPN networks for secure access.
- Cloud - AWS, Azure, multi-cloud architectures
- On-premise - Bare metal servers, Docker Swarm, network architecture
- Databases - PostgreSQL, MySQL, Redis, vector databases (pgvector, Pinecone)
- Monitoring - Grafana, VictoriaMetrics, Zabbix, alerting pipelines
- Security - VPN, firewall configuration, TDE encryption, access control
Frequently Asked Questions
What is a fractional CTO and how does it work?
A fractional CTO provides senior technical leadership on a part-time or project basis. At Happyberg Labs, this typically means 2-3 days per week of hands-on involvement: setting technical direction, mentoring your engineering team, making architecture decisions, and managing technical debt. It gives startups and SMEs access to 20+ years of CTO-level experience without the full-time cost, which can save 60-70% compared to a full-time senior hire.
What AI and machine learning services do you offer?
We build production AI systems including agentic AI workflows with LangChain and LangGraph, computer vision pipelines with MediaPipe and PyTorch, NLP and embedding systems with vector databases, and custom ML models. We focus on practical AI that solves real business problems. Our founder has shipped AI systems in digital health, lab automation, and enterprise operations.
Do you work with healthcare and digital health projects?
Yes, digital health is a core specialization. Our founder has co-founded two digital health companies and has deep experience with DICOM medical imaging, HL7 and FHIR interoperability standards, telemedicine platforms, EHR integration, and clinical workflow automation. We have shipped telemedicine systems connecting facilities across multiple continents.
Where are you based and what regions do you serve?
Happyberg Labs is based in Sada, A Coruña, Galicia, Spain. We serve clients worldwide, working remotely across time zones. Our founder has worked with teams in Europe, North America, South America, and Antarctica. We communicate in English and Spanish.
What technology stack do you use?
We choose technology pragmatically based on the problem. Primary languages: Python (AI/ML, data, automation), Elixir (real-time, distributed systems), and Ruby (web applications, APIs). Infrastructure: PostgreSQL, Docker, Linux, AWS, Azure, and bare metal. AI stack: LangChain, LangGraph, PyTorch, FastAI, MediaPipe, and vector databases like Pinecone and pgvector.
How do you approach product development?
We follow pragmatic product methodologies: Shape Up for scoping and shipping in 6-week cycles, Jobs to Be Done for understanding user needs, OKRs for aligning work with outcomes, and YAGNI to avoid over-engineering. We ship something every Friday. We believe in boring technology that works, small teams with high trust, and building the minimum needed to validate and iterate.