I am a Senior Solution Architect
30+ years making data fast, secure and intelligent – from Oracle, PostgreSQL and MySQL clusters and scalable AWS architectures to fine-tuned language models and self-improving AI agents.
Hands-on expertise across the full data stack – the databases that run your business, the Linux and cloud platforms underneath them, and the AI that turns that data into decisions.
Installation, upgrades, patching, security, backup & recovery and remote support for Oracle, SQL Server, PostgreSQL, MySQL, DB2, Sybase and SQLite, plus data modelling and SQL, PL/SQL and T-SQL development.
Find out why your database is slow. Query and index tuning, configuration review and capacity planning, delivered as a clear report with prioritised fixes.
MySQL Cluster, replication, failover and disaster recovery. Low-risk migrations between database engines and to AWS (RDS for MySQL, PostgreSQL and Oracle, DynamoDB, Redis).
AI agents that monitor your databases around the clock, find anomalies early, explain the root cause in plain language and recommend fixes before your users notice.
Small and large language models fine-tuned on your own data. Private, cost-efficient models built for your domain, with evaluation to show they actually work.
Recursive self-improving solutions that learn from every new data point. One example is SPAM Escudo, which keeps getting better at stopping spam and fraudulent sign-ups.
MIT-certified data science: data preparation, topic modelling, predictive models and statistics that turn raw data into decisions.
Secure, scalable, highly available AWS applications: serverless back ends with API Gateway, Lambda and event triggers, Cognito authentication, CloudFront and S3 delivery, Application and Network Load Balancers with Auto Scaling Groups, and RDS for MySQL, PostgreSQL and Oracle. Every release goes through CI/CD pipelines with safe, fast rollback.
Certified Linux administration, hardening and automation, plus custom applications in Python, C, C++, C#/.NET, Go and Rust. AI-assisted code reviews with improvement recommendations, in almost any language.
A fixed-scope review of your database or platform, with a written report and an action plan.
Cloud architectures, migrations, HA builds, AI agents and model fine-tuning, delivered on time and to spec.
Remote DBA and Linux support on retainer, for clients across Europe, Asia-Pacific and the USA.
I am an internationally experienced Senior Data Solution Architect. I started out as an Oracle DBA and went on to work with the major relational platforms (Oracle, SQL Server, PostgreSQL, MySQL, DB2 and Sybase), NoSQL and in-memory stores such as DynamoDB and Redis, Linux/UNIX, networking and AWS cloud architecture. From there I moved into big data, machine learning and, most recently, applied AI.
In 2006 I founded IT Integrated Business Solutions, delivering database, Linux and cloud projects remotely for clients across Asia-Pacific, Europe and the USA. As Senior Data Solution Architect for Runner Technologies, I designed the database and cloud architecture for CLEAN_Data Portal and CLEAN_Update. Both are .NET applications built for high availability on AWS, running on EC2 Auto Scaling groups (behind an Application Load Balancer for CLEAN_Data Portal) with RDS MySQL replicated across Availability Zones for failover. Both were penetration-tested with the development teams.
My current focus is where databases and AI meet: AI agents that keep databases healthy, fine-tuned small and large language models, and recursive self-improving systems such as SPAM Escudo. I bring a logical, analytical approach to complex problems and take ideas all the way from concept to production.
With AI-assisted tooling I can also read code in other languages, review it and recommend improvements.
Each of these products runs on an AWS architecture I designed, with CI/CD pipelines for automated, repeatable releases and fast rollback if a deployment fails.
An identity-verification and anti-fraud API for Higher Education, aligned with NIST IAL2, that helps stop FAFSA fraud and “ghost students”. It cross-references applicant names, addresses, phone numbers, emails, SSNs and IP addresses, and returns a confidence score with fraud analysis, in real time or in batch. Cognito secures access to both the API and the customer portal.
An AI-powered spam classifier for contact forms. Each submission passes through a five-phase pipeline: honeypot and cryptographic bot detection, schema checks, a rules engine that responds in under 50 ms, multilingual semantic embeddings in S3 Vectors, and LLM analysis on Amazon Bedrock for truly ambiguous messages. It is recursively self-improving, so detection keeps getting sharper while real leads still get through. Sites add it with a JavaScript snippet or a WordPress plugin.
A software studio platform and product family. It includes SPAM Escudo and Transit Ledger, an invoice and receipt processor for Google Workspace that extracts line items, splits taxes and files records to Drive and Sheets without keeping any data in a database. An on-premise edition with offline models is in development.
A self-service portal where users upload a file of addresses and have them validated against an internal address database using Runner EDQ's CLEAN_Address. I designed the database and the highly available architecture. An Application Load Balancer routes traffic to EC2 target groups, scaled by Auto Scaling, running a .NET application which parses each file and forwards the address data to the validation server. RDS MySQL with automatic failover sits behind it. The system was penetration-tested with the development team.
A .NET application that delivers data and software updates for Runner's address-verification products. It runs on an EC2 Auto Scaling group, with RDS MySQL replicated to a second Availability Zone so the service survives the loss of an AZ.
Autonomous agents that watch database health metrics, find emerging problems, diagnose the likely root cause and recommend fixes. They combine decades of DBA know-how with modern AI.
Used topic modelling to uncover hidden themes in customer support tickets. This led to process improvements that significantly reduced the number of tickets raised and improved the customer experience.
A machine-learning spam filter for a WordPress contact form. Registration data was enriched with location features taken from IP addresses and phone numbers (country, state/city) to flag suspicious sign-ups.
An AWS CLI utility for CLEAN_Cloud that discovers servers, drains each one from its load-balancer target group, updates data and software, then restarts and re-registers it, with disaster-recovery integration built in.
A Linux and Windows datafile and software updater using OAuth2 authentication, AWS signed-URL downloads and licence-key entitlements. It significantly cut update-related support tickets.
A contact-form plugin with spam filtering and Keap (InfusionSoft) CRM integration that removes duplicate contacts by email, giving marketing and sales teams cleaner leads.
Developed and marketed more than 100 UI extensions for the Mac web-design tool RapidWeaver, in Objective-C with JavaScript, CSS and HTML. They are used by web developers worldwide.
Got a slow database, a migration coming up or an AI idea you'd like to test? Tell me about it using the form below and I'll get back to you.
San Sebastián, Spain
Working remotely worldwide
English
Español
Français