AI Product Engineering
LLM workflows, AI-assisted tools, MCP integrations, local model experiments, custom memory concepts, and reasoning systems.
Axefield
Available for full-time and contract roles
I am Pedro "Axefield" de Castro Davila, a TypeScript-first systems builder connecting LLM tooling, Node.js platforms, automation workflows, dashboards, UX-centered interfaces, and creative technology into usable products.
What I build
The through-line is turning ambiguous technical possibilities into working interfaces, workflows, and platforms.
LLM workflows, AI-assisted tools, MCP integrations, local model experiments, custom memory concepts, and reasoning systems.
TypeScript, Node.js, Express services, REST APIs, dashboards, frontend interfaces, auth flows, and SaaS-style tools.
User flows, interface clarity, design implementation, conversion paths, performance, accessibility, and product polish.
Workflow automation, trading dashboard concepts, data pipelines, signal surfaces, internal tools, and operational systems.
Unreal Engine experiments, interactive worlds, narrative IP systems, AI media generation, and game-adjacent production tools.
WordPress, Shopify, CMS/CRM/SaaS management, technical SEO, page speed, backups, integrations, and client-ready delivery.
Proof map
Each case study is framed so a hiring team can see the problem, architecture, tradeoffs, quality bar, deployment path, and product value behind the code.
Self-directed AI/product engineering lab covering subdomain apps, auth, dashboards, email systems, learning modules, and UX architecture.
C++ AI systemsCMake-based C++ decision/inference system with mythic headers, server components, protobuf/gRPC examples, shaders, and AI orchestration design.
AI systemsModel Context Protocol integrations, local LLM workflows, reasoning tools, Cursor/Claude/ChatGPT workflows, and fine-tuning exploration.
AutomationPersonal trading dashboard concepts, signal workflows, market analysis surfaces, and automation systems for decision support.
Creative techGame-adjacent systems, Unreal Engine exploration, narrative IP, AI-assisted media workflows, and worldbuilding tools.
Engineering evidence
I want interviewers to find the concrete work quickly: languages mapped to systems, systems mapped to technical problems, and technical problems mapped to delivery outcomes.
Built the Truligon platform surface: Express routing, subdomain handling, admin/auth flows, dashboard concepts, Vite build outputs, and production-facing static delivery.
Used Python for Mythogon configuration, research, PDF/protection utilities, and AI workflow support where repeatable scripts reduce manual setup and validation work.
Worked with Prisma, SQLite, schema-backed services, dashboard data concepts, and reasoning or memory stores that need clear data boundaries and maintainable access paths.
Developed Mythogon as a CMake-based C++ system with decision-engine headers, server modules, protobuf/gRPC examples, shader experiments, and performance-oriented architecture notes.
Use TypeScript checks, frontend builds, generated-page parse fixes, accessibility scans, code review habits, and explicit route verification to keep changes maintainable.
Maintain Cloudflare tunnel routing, subdomain configs, static asset paths, canonical pages, and production recovery steps so shipped surfaces stay reachable and debuggable.
Company fit
I can connect LLM capabilities to usable product surfaces, tool workflows, and internal automation.
I can prototype across frontend, backend, APIs, dashboards, onboarding flows, and product UX without waiting on perfect specs.
I bring practical delivery experience across WordPress, Shopify, technical SEO, page speed, CMS operations, and client-facing UX.
I bridge systems design, AI tools, Unreal workflows, storytelling, interactive media, and production tooling.
Hiring packet
I am currently focused on roles in AI engineering, full-stack TypeScript development, LLM tooling, UX engineering, automation, and creative technology.