Senior technology consultancy

Intelligent systemsSoftware, data and models for real-world challenges.

We design and build software, data platforms, artificial intelligence and mathematical models for organisations that need reliable systems ready for production.

What we do
0+
Years of experience
0+
Areas of expertise
3
Cloud platforms

Founder's professional background

Nokia
BCC
What we do

Complex problems.Clear, robust systems.

We work on projects where software, data, models and infrastructure form a single system. Before building, we establish the objective, the users, the constraints, the risks and the operating environment.

Data engineering

Reliable, well-governed data platforms

We bring data sources, business rules and operational workflows into a coherent architecture. Quality, lineage, ownership and cost are built into the design rather than added later.

Lakehouse architectures · regulated environments · global supply chains

Artificial intelligence

AI embedded in real workflows

A model creates value only when it works with the right data, permissions, tools and people. We design evaluation, latency, cost controls and failure handling as part of the system.

Production LLM systems · document retrieval · enterprise integration

Machine learning

Predictive models built for production

We design the full lifecycle: feature engineering, training, validation, deployment and monitoring. Each model is tied to a decision and an accountable owner.

Risk models · scientific research · production monitoring

Mathematical optimisation

Better decisions under complex constraints

We model planning, allocation, routing and scheduling around the constraints that exist in practice. Mathematical optimisation, simulation and software produce decisions people can understand and use.

Routing · resource allocation · planning under multiple constraints

Scientific computing

Reproducible scientific software

We turn experimental workflows into maintainable software without compromising scientific rigour. Data, parameters and results remain traceable, repeatable and ready for further analysis.

Genomics platforms · oncology data · published biomedical research

Custom software

Internal tools that improve operations

We build portals, applications and services around the way each organisation actually works. Product, backend and data are designed together to simplify processes and remove manual work.

Operational portals · public-sector processes · backend and data services

Experience

More than 20 years in demanding environments

Selected work from the founder's career across enterprise platforms, regulated banking, consumer goods, global digital services, genomics and oncology research. Different sectors, one consistent requirement: reliable technology that works in practice.

Selected experience

01/05

Global consumer goods

Cloud governance and architecture for a global organisation

Cloud architecture

Led the technical design of an Azure platform for a global consumer goods group, aligning internal teams, vendors and delivery partners around a shared architecture, operating model and Terraform-managed infrastructure.

Azure governanceTerraform infrastructureVendor alignmentOperating standards

01/05

BCC · IBM / Viewnext

An enterprise data platform built from the ground up

Data engineering

Designed and built BCC's first large-scale data platform: a data lake, real-time Kafka pipelines and an operational data journal that gave technical and business teams a clear view of system state and data movement.

Data lake from scratchReal-time KafkaCross-team adoptionTechnical training

02/05

Nokia · Microsoft MixRadio

Recommendation systems for a global music service

Recommendation systems

Built the AWS data layer behind MixRadio's recommendation service, turning data science algorithms into production personalisation for millions of users. The work also included APIs, microservices and internal catalogue tools for editorial teams.

Recommendation pipelinesAWS in productionMillions of usersInternal tools

03/05

Coral Genomics

Petabyte-scale genomics for machine learning

AI for genomics

Designed and released DNARecords, a sparse genomics format and open-source SDK that transforms large VCF/BGEN datasets into efficient representations for machine learning and deep learning.

bioRxiv publicationOpen-source SDKVCF/BGEN conversionGenomics infrastructure

Oncko

Software and AI for oncology research

AI for oncology

Built scientific software and data systems for drug-combination research: large-scale matrix clustering, bioinformatics pipeline orchestration, LLM-assisted extraction, data harmonisation and hypothesis tracking.

Drug combinationsBioinformatics pipelinesLLM extractionHypothesis management

05/05

Global consumer goods

Cloud governance and architecture for a global organisation

Cloud architecture

Led the technical design of an Azure platform for a global consumer goods group, aligning internal teams, vendors and delivery partners around a shared architecture, operating model and Terraform-managed infrastructure.

Azure governanceTerraform infrastructureVendor alignmentOperating standards

01/05

BCC · IBM / Viewnext

An enterprise data platform built from the ground up

Data engineering

Designed and built BCC's first large-scale data platform: a data lake, real-time Kafka pipelines and an operational data journal that gave technical and business teams a clear view of system state and data movement.

Data lake from scratchReal-time KafkaCross-team adoptionTechnical training

02/05

Nokia · Microsoft MixRadio

Recommendation systems for a global music service

Recommendation systems

Built the AWS data layer behind MixRadio's recommendation service, turning data science algorithms into production personalisation for millions of users. The work also included APIs, microservices and internal catalogue tools for editorial teams.

Recommendation pipelinesAWS in productionMillions of usersInternal tools

03/05

Coral Genomics

Petabyte-scale genomics for machine learning

AI for genomics

Designed and released DNARecords, a sparse genomics format and open-source SDK that transforms large VCF/BGEN datasets into efficient representations for machine learning and deep learning.

bioRxiv publicationOpen-source SDKVCF/BGEN conversionGenomics infrastructure

Oncko

Software and AI for oncology research

AI for oncology

Built scientific software and data systems for drug-combination research: large-scale matrix clustering, bioinformatics pipeline orchestration, LLM-assisted extraction, data harmonisation and hypothesis tracking.

Drug combinationsBioinformatics pipelinesLLM extractionHypothesis management

05/05

Global consumer goods

Cloud governance and architecture for a global organisation

Cloud architecture

Led the technical design of an Azure platform for a global consumer goods group, aligning internal teams, vendors and delivery partners around a shared architecture, operating model and Terraform-managed infrastructure.

Azure governanceTerraform infrastructureVendor alignmentOperating standards

01/05

BCC · IBM / Viewnext

An enterprise data platform built from the ground up

Data engineering

Designed and built BCC's first large-scale data platform: a data lake, real-time Kafka pipelines and an operational data journal that gave technical and business teams a clear view of system state and data movement.

Data lake from scratchReal-time KafkaCross-team adoptionTechnical training

02/05

Nokia · Microsoft MixRadio

Recommendation systems for a global music service

Recommendation systems

Built the AWS data layer behind MixRadio's recommendation service, turning data science algorithms into production personalisation for millions of users. The work also included APIs, microservices and internal catalogue tools for editorial teams.

Recommendation pipelinesAWS in productionMillions of usersInternal tools

03/05

Coral Genomics

Petabyte-scale genomics for machine learning

AI for genomics

Designed and released DNARecords, a sparse genomics format and open-source SDK that transforms large VCF/BGEN datasets into efficient representations for machine learning and deep learning.

bioRxiv publicationOpen-source SDKVCF/BGEN conversionGenomics infrastructure

Oncko

Software and AI for oncology research

AI for oncology

Built scientific software and data systems for drug-combination research: large-scale matrix clustering, bioinformatics pipeline orchestration, LLM-assisted extraction, data harmonisation and hypothesis tracking.

Drug combinationsBioinformatics pipelinesLLM extractionHypothesis management

05/05

How we work

From complexity to a working system

Every engagement starts with a concrete outcome: a platform, an application, a model in production, an architecture decision or a migration. We work closely with the people who understand the problem and will ultimately operate the solution.

01

Understand the real problem

We examine the current system, its data, users, constraints and risks. The aim is to define the decision that must be made and the outcome the work must achieve.

02

Design before building

We set out the options, trade-offs and risks. The architecture is reasoned through before significant time and budget are committed to the critical parts.

03

Build the core of the solution

Implementation focuses on what creates value and makes production possible: data, infrastructure, logic, interfaces and operational behaviour.

04

Leave the team in control

We leave repositories, infrastructure, operational documentation and decision records in a state the team can maintain and evolve independently.

Engineering judgement

Rigour in every decision

Tools change; sound engineering does not. What matters is understanding the constraints, explaining the choices and keeping the system clear as it grows.

Design for failure

Data quality, latency, permissions, retries and drift are considered from the outset, not after problems appear.

Decisions have a rationale

Every architectural choice is explained with its alternatives, consequences and fit with the client context.

The client stays in control

Code, infrastructure, documentation and knowledge are part of the delivery and remain with the client team.

Purposeful simplicity

We favour readable systems, direct interfaces and predictable operations over complexity that adds no value.

Sunny Data

Senior expertise
from start to finish

Sunny Data is an independent technology consultancy based in Spain. Every engagement is led by the founder, who remains directly involved in discovery, architecture, implementation and delivery.

We combine software and data engineering, artificial intelligence, mathematical optimisation and scientific computing. This breadth allows us to address the whole system rather than one isolated component.

Our way of working is direct and productive: understand the problem, make well-founded decisions, build what is needed and leave behind a clear, maintainable solution.

Professional background

Mathematician · Senior technologist

More than 20 years designing, building and running systems across telecommunications, entertainment, consumer goods, finance and biomedical research.

IBM · Nokia · Microsoft · BCC

Expertise

The sector changes, but the need is consistent: bring software, data, models and infrastructure together in a system that works and can evolve.

Cloud and data platforms
Technical foundations, governance and operating models that allow multiple teams to work without losing control.
Production data engineering
Real-time processing, data lakes and architectures designed for reliability, traceability and daily use.
AI and mathematical models
Recommendation systems, predictive models, optimisation and LLM solutions integrated into operations.
Scientific software and infrastructure
Genomics, oncology, bioinformatics pipelines and reproducible systems for research with heterogeneous data.
20+ years
Technical experience
Mathematics + software
Core disciplines
New projects

When the problem demands
judgement and delivery

We turn complex technical problems into clear decisions and working systems. Tell us what you need to solve.