Foundation Skills
Systems thinking, languages, data and the fundamentals beneath every generation of technology.
JOHN DUNNINGAI SYSTEMS · SOFTWARE ENGINEERING · FULL STACK ENGINEERING · BUSINESS RULE MANAGEMENT SYSTEMSUK / 2026
JDJOHN DUNNINGENGINEERING × INTELLIGENCE
SECURE MY SKILLS ↗
serious software
for the AI era.
Over two decades of engineering experience — from secure operating systems and global payments to cloud platforms, deterministic AI guardrails and agentic knowledge systems.
JDSYSTEMS
AIJAVARAG
AWSRULESKAFKA
RUSTSECUREAGENTS
ENGINEERING×INTELLIGENCE
Informix / C→Secure AIX→Java→Global Payments→Rules Engines→Cloud + AI→Agentic Systems→Spring Boot→Kafka / Services→RAG→Deterministic Guardrails→
Informix / C→Secure AIX→Java→Global Payments→Rules Engines→Cloud + AI→Agentic Systems→Spring Boot→Kafka / Services→RAG→Deterministic Guardrails→
SYSTEMS EVOLUTION
A legacy of enterprise solutions.
Security & Trust
Secure systems, controlled boundaries and engineering people can depend on.
Maintainability
Software designed to evolve, survive change and remain understandable long after delivery.
Integrity
Auditable decisions, deterministic rules and systems whose behaviour can be explained.
Intelligence
AI, RAG and agents added to disciplined engineering rather than substituted for it.
SELECTED STORIES
Three projects.Three reasons to call.
Great portfolios do not bury the strongest work in twenty cards. These are the stories that best explain how I operate: ambiguous problem, difficult constraints, engineered outcome.
01 / AI SYSTEMS
CURRENT · 2026
Documents becomeconnected knowledge.
Millions of files are useless if the knowledge inside them cannot be found, connected or trusted.
Problem Fragmented documents, email and events across large personal and professional archives.
Build Modular ingestion, deduplication, extraction, embeddings, timelines and AI-assisted retrieval.
Stack Tauri · Rust · React/Vite · SQLite · vector search · local/cloud LLMs.
PDFMAILDOCIMG
QUERY“What happened, when — and why?”Evidence-linked answer →
02 / EXPLAINABLE AI
AI × DETERMINISM
AI that can think,rules that stay in control.
Enterprise AI needs more than a good prompt. It needs boundaries that can be tested, audited and explained.
Pattern LangChain4j and LLM workflows paired with Drools decision layers.
Benefit Probabilistic intelligence inside deterministic policy and business constraints.
Experience Rules architecture across payments, insurance, government and defence.
LLM proposes action confidence 0.94
POLICY GATERULE 421
✓
ALLOWEDdecision trace retained
03 / DELIVERY
NORTHROP GRUMMAN · 2023–2026
Turn difficult researchinto working reality.
Technical leadership is not a title. It is what happens when the difficult project has to work.
Context Secure-cloud, cyber and AI delivery in a DevSecOps environment.
Work Java, Spring Boot, AWS, Python, technical leadership and mentoring.
Outcome Research programme recovered and demonstrated to senior defence stakeholders.
MISSION SYSTEMDEMO READY
THE TOOLKIT BEHIND THE WORK
Modern AI.
Embracing the future of information technology.
LLMs are powerful. Enterprise systems still need determinism, auditability and engineering discipline. My current work combines both.
CURRENT WORK · 2026
Document intelligence→structured knowledge
Automated pipelines ingest emails, documents and attachments, extract entities and events, build chronologies and make the result searchable through local and cloud AI.
- Tauri + Rust desktop services
- React / Vite interfaces
- SQLite + vector search
- Local & cloud LLM integration
- Code-enforced AI guardrails
01Documentsemail · PDF · files
02Ingestindex · dedupe · parse
03Understandentities · events · places
04KnowledgeRAG · timeline · search
LLM+RULES✓
CONTROLLED AI
LLMs with deterministic guardrails
LangChain4j and Drools combine probabilistic AI with auditable business rules, keeping critical decisions explainable and controlled.
Kafka / Services
EVENT-DRIVEN
Systems built for real traffic
Spring Boot, Kafka and Kafka Streams supporting national-scale logistics and high-volume transactional workloads.
SC
secure_cloud://readySECURE DELIVERY
Defence, cyber and secure cloud
Led AI, cyber and DevSecOps initiatives at Northrop Grumman, including research demonstrated to the Deputy Defence Minister.
Selected evidence
Not a technology list.A record of delivery.
Global card processing
Created and modified JRules rule sets supporting Visa Europe transaction processing.
European payments regulation
Designed Discover's real-time and batch rules system; delivered ahead of schedule, under budget and with very low live defects.
Research turnaround
Rescued a failing Northrop Grumman research programme and delivered live demonstrations to senior defence stakeholders.
Secure operating systems
Worked on IBM RAF LITS, Shield and B1 Secure AIX kernel engineering, device drivers and secure OS design.
Rules architecture
Architected governed Drools platforms across insurance, government, payments and enterprise systems.
Built from first principles
Career work includes a database engine, JDBC driver, XML database, device-driver engineering and a web server built from scratch.
Experience
From systems programming to AI systems architecture.
Recent roles are expanded. Earlier career remains accessible without turning the page into a wall of text.
2026 — NOW
Independent Software Developer AI & Document Intelligence Systems
Designing modular document management, case-timeline and event-management applications with Tauri, Rust, SQLite, React/Vite, local/cloud LLMs and code-enforced guardrails. Building automated ingestion, indexing, entity/event extraction and chronology systems, plus companion iPhone applications.
AIRustTauriReactRAGLocal LLMs
2023 — 2026
Lead Software Engineer / Project Manager Northrop Grumman UK · Mission Systems
Led AI, cyber and secure-cloud DevSecOps initiatives using Spring Boot, AWS, Java and Python. Integrated LangChain4j with Drools for explainable decisioning, mentored engineers and rescued a research programme later demonstrated to the Deputy Defence Minister.
JavaSpring BootAWSPythonLangChain4jDrools
2021 — 2023
Full Stack Developer Yodel
High-throughput Spring Boot services with Kafka/Kafka Streams, MongoDB, Informix, Elasticsearch and Kibana; React Native handheld and React/Redux office systems.
2015 — 2019
Drools Architect Level Global · Swinton · Charity Commission · Discover
Rules architecture, Oracle ODM/OPA migration, AWS deployment, quotation/claims logic, government Live Beta delivery and European payment-processing regulation.
2007 — 2015
Enterprise Java & Rules Visa · MBNA/Bank of America · EBRD · Scottish Government · Atos
Global card transactions, credit-card integration, financial risk/exposure, BRMS/BPM, complex event processing and enterprise integration.
1988 — 2007Earlier engineering careerExpand ↓
IBM RAF LITS / Security — AIX device drivers, lexical analysers, Shield and B1 Secure AIX kernel work.
IBM Global Services — worldwide contract-tracking intranet, Java/J2EE, WebSphere and DB2.
HP / T++ — senior Java/J2EE/C trainer; founder-member era at T++.
BT Research Labs — deployment/project management across UK teams with a £0.5m budget.
MOD / ICL / PSA / HMRC — defence, banking, government and transaction-processing systems.
Early commercial systems — C, Informix, Unix/Xenix, database and financial systems from 1988 onward.
Capability map
Deep engineering, not keyword collecting.
AI / Knowledge
Agentic AI · RAG · GraphRAG · embeddings · vector search · local LLMs · LangChain4j · deterministic guardrails · document intelligence
Java / Backend
Java 8–17 · Spring Boot · Spring MVC · Spring Batch · Hibernate · JEE/J2EE · REST · microservices
Rules / Decisioning
Drools 5/6 · Drools BRMS · JRules · DRL · BPMN · decision tables · complex event processing · explainable logic
Messaging / Integration
Kafka · Kafka Streams · JMS · MQ · RabbitMQ · Apache Camel · CXF · SOAP/REST · XML · JSON
Cloud / Delivery
AWS · Lambda · S3 · DynamoDB · Route53 · Docker · OpenShift · Jenkins · CI/CD · Git · Maven · Gradle
Data / Search
Oracle · PL/SQL · PostgreSQL · MongoDB · SQLite · Informix · DB2 · Elasticsearch · Kibana · high-volume transaction design
Recommendations
People I’ve worked with.Feedback
LinkedIn recommendations from colleagues, clients and managers — in their own words.
"
John has done some outstanding work for me and he has worked extremely hard. I would recommend him to anyone.
"
John is a very knowledgeable, conscientious professional who always delivers first rate results within the required timescales. I would have no hesitation in recommending him.
"
Working with John has been enjoyable and I found John to be knowledgeable, professional and someone who both works well to deliver under high pressure and adapts quickly to changing requirements. I would be more than happy to work with John in the future.
"
Even though John joined the project at a difficult point, he showed great expertise and contributed with a raft of innovative ideas to help the team. He was a very popular and valued member of the team.
"
John has very strong Java knowledge and delivered what was required from him to a high quality standard. Further John exceeded expectations in providing mentoring and help to less experienced developers.
"
Absolutely excellent! Covered all the topics needed prior to the certification tests. John was a knowledgeable and interesting presenter, able to involve all the attendees and allowed everyone to get the utmost from the experience.
"
John provided a two-day Java training workshop on a one-to-one basis. I was very happy with the format, content and delivery. The course provided an invaluable foundation for my Java skills — I went on to attain Sun Java accreditation.
"
I was very satisfied with having John on my team, especially because of his positive attitude, creativity, eagerness to learn, dedication and professionalism.
Let's build something difficult.
Enterprise engineering.
Modern AI.
Available for senior engineering, architecture and technical leadership opportunities.