Joshua Fields — Software Engineer & Creative Technologist

Monstro — Senior Full Stack Engineer (Mar 2026 – Present)

Monstro was building AI-driven infrastructure for financial and institutional workflows, combining assistant-style interfaces, authorization systems, data pipelines, internal tooling, and cloud-native backend services. The product operated in a technically demanding environment where permissions, data ownership, institution workflows, and AI-assisted user experiences all needed to work reliably across multiple services.

Joshua worked at Monstro as a Senior Full Stack Engineer, contributing across backend architecture, authorization, AI assistant infrastructure, CMS integration, financial data modelling, and cross-repository platform delivery. His work spanned six repositories, including services, SpiceDB, copilot, institution portal, data gateway, and shared CI.

One of Joshua's key contributions was designing and shipping Monstro's SpiceDB authorization foundation end-to-end. This included Kubernetes manifests, schema-as-code, a 12-role by 70-capability permissions matrix, a Go client library, and a shadow-mode rollout across production handlers. The system created a Zanzibar-style authorization layer that other engineers could build on daily, giving the platform a more scalable and maintainable approach to access control.

Joshua also led a backend-wide subject-key migration from customer-based access to segment-based access. This changed how permissions and data access were gated across the platform, unblocking important product areas including the jobs engine and the consumer-to-banker messaging rails.

Beyond authorization, Joshua owned the copilot-to-services-to-CMS integration layer. He helped move block-grain ownership into the services tier and shipped a field-registry-aware classification pipeline, which was later consumed by the institution portal. This work strengthened the connection between Monstro's AI assistant experience, backend services, and institution-facing workflows.

Joshua shipped several major MVP epics, including role-based access, configurable role permissions, and book management. He also drove the services-side Vault MVP across a broad financial holdings surface, covering real estate, insurance, cash and savings, investments, vehicles, mortgages, credit-card liabilities, and institution views.

As part of this work, Joshua built the L1/L2/L3 holdings API using a typed-column model and integrated third-party enrichment providers such as AbodeMine for address classification and VinAudit for vehicle valuation. These enriched holdings fed into Monstro's net-worth pipeline through Kafka data_points events, connecting product-facing financial data to downstream analytics and intelligence workflows.

Across his time at Monstro, Joshua delivered 50+ tickets, hardened the Go/SSE AI assistant, filed and fixed cross-cutting platform bugs, and regularly ran review sweeps to unblock teammates. His work demonstrated the ability to operate at platform level: designing core systems, shipping product-facing features, improving developer workflows, and contributing across infrastructure, backend services, AI orchestration, and financial data products.