AI Agents & LLM Integrations
RAG pipelines, agentic workflows, prompt engineering, and intelligent tooling — from prototype to production under enterprise governance.
RAGAI AgentsLLMPrompt EngineeringMCP
Our expertise
High-Scale Architecture
Event-driven systems, multi-tenant SaaS, real-time pub/sub, and performance under concurrency — designed for resilience, not slide decks.
MicroservicesPostgreSQLLISTEN/NOTIFYRedisDocker
Systems Integration & Legacy Modernisation
Complex API design, ecosystem integrations, and modernising legacy platforms — bridging stakeholders and engineering teams.
C# / .NETAPIsiMISSolution Design
Our process
- 1 Consult Understand the problem, constraints, stakeholders, and what success actually looks like — including what is not written down.
- 2 Architect Design a pragmatic solution — right-sized architecture, clear boundaries, and a path that teams can own after handover.
- 3 Build Implement, integrate, test, and document. Hands-on delivery or technical leadership through your existing team.
- 4 Handover Transfer knowledge, operational runbooks, and systems your team can run — or stay engaged as principal architect through delivery.
How we engage
Typical path: introductory call → scoped proposal → engagement (time & materials or fixed scope).
- 1 Architecture & discovery Review existing systems, define target architecture, and deliver written recommendations.
- 2 Hands-on build End-to-end delivery or augmentation of your in-house team.
- 3 Technical leadership Embed as principal architect through delivery — mentoring included.
- 4 AI agent POCs Scoped prototypes for RAG, agents, or workflow automation.
Selected work & experience
Work from Matthew Heffernan's career — employment and principal-led product engineering. Representative of capability; not implied client engagements unless stated.
Employed role · ABC Innovation Lab
Enterprise Media — AI Systems
As Senior Software Engineer at the ABC Innovation Lab: deployed a cross-system AI Chrome extension that unified siloed enterprise tools into one experimentation harness — rapid newsroom adoption and an estimated ~6 months of sprint time saved. Built RAG pipelines for video metadata and delivered Video Chapters with embedded quizzes used nationwide in schools.
RAGLLMChrome ExtensionReactTypeScriptAzure/GCP
PolymorphicIT product engineering · B2B2C SaaS
Multi-Tenant Booking Platform — Booka.work
Architected a high-concurrency B2B2C scheduling platform for salons and solo professionals. Request-scoped tenant isolation on PostgreSQL, real-time calendar sync via LISTEN/NOTIFY, Redis rate limiting, Stripe Connect, Playwright E2E and k6 load testing.
Next.jsPostgreSQLDrizzleLISTEN/NOTIFYMulti-tenantStripe
Employed role · CauseIS · iMIS
NFP Ecosystem Integration
As Senior Solutions Consultant: delivered complex API integrations and microservices for not-for-profit clients on the iMIS platform. Technical anchor for the team — enabling developers to take on integrations beyond their previous comfort zone.
C#APIsiMISMicroservicesIntegration
Leadership
Matthew Heffernan
Principal Systems Architect & Founder
Computer scientist with 20+ years bridging technical complexity and business reality. Matthew provides the strategic direction and safety net that lets teams tackle high-stakes challenges — from national broadcaster AI systems to multi-tenant SaaS and real-time game infrastructure.
Parent company of Split Personas game studio. Based in the Blue Mountains, NSW — hybrid and remote engagements across Australia.
Contact
Contract and architecture engagements welcome. Tell us what you are building — we will tell you honestly if we are the right fit. References available on request.
info@polymorphicit.com.au · +61 466 954 545
Start a conversation