Systems Engineer & AI Architect
Regulated platforms • Agentic AI • Rust/TypeScript monorepos • DeFi & regenerative agriculture

I work where AI, legacy systems, and real-world risk collide:

• Agent/LLM platforms with governance, cost & telemetry baked in
• DeFi / chain systems that have actual risk math, not just vibes
• Farm & land systems where business plans, software, and soil line up
• Delivery / IaC / PR automation that makes all of this repeatable

If you’re skimming, start with:
cognitive-mesh, CognitiveMeshUI,
autopr-engine,
PhoenixRooivalk,
farm-business-plan, zeeplan, pigpro,
vv-docs, vv-landing.

• Cognitive Mesh & human–AI collaboration
  Orchestrating tools/agents with RBAC, audit, policy-as-code, and observability. Built to survive regulated environments, not just “demo calls”.

• VeritasVault – DeFi risk & chain governance
  Chain contracts, services, landing, and docs wired together via docs + IaC so the whole surface is governable and explainable. :contentReference[oaicite:0]{index=0}

• PhoenixRooivalk – Rust + TS counter‑drone platform
  A serious monorepo (Rust + TypeScript + Next.js + Docusaurus, Turborepo) for a c‑UAS evidence & control stack with dual-chain anchoring. :contentReference[oaicite:1]{index=1}

• Regenerative agriculture & “skin‑in‑the‑game” design
  Multi-year farm plans, financial models, and working software aimed at cash flows, not dribbble screenshots. :contentReference[oaicite:2]{index=2}

• Delivery, infra & governance
  IaC, policy gates, what‑if validation, PR automation, and template repos that bake in tests, type-checking, and CI.

Repo: https://github.com/JustAGhosT/cognitive-mesh :contentReference[oaicite:3]{index=3}

• Stack:
  • Backend: C# / .NET 9 service
  • Targets: Azure OpenAI, RAG integrations
  • Surroundings: RBAC, audit logging, policy-as-code, cost/usage observability
• What it shows about me:
  • I treat “agents” as production services: layered architecture, least-privilege, and explicit governance instead of ad-hoc scripts.
  • Hooks for eval and telemetry, not just “call the model and hope”.

Repo: https://github.com/JustAGhosT/CognitiveMeshUI :contentReference[oaicite:4]{index=4}

• Stack:
  • Next.js + React + TypeScript
  • Tailwind, Storybook, Style Dictionary
• Focus:
  • Token-driven component library
  • Screens for agents, tools, audit events, and cost telemetry
• Value: Front-end architecture that treats AI as infrastructure: observable, inspectable, and explainable.

Repo: https://github.com/JustAGhosT/autopr-engine :contentReference[oaicite:5]{index=5}

• Stack:
  • Core: Python 3.8+ package (published to PyPI)
  • Supports: Docker, GitHub Action / Marketplace app
  • Extras: Small TypeScript + PowerShell utilities
• What it does:
  • Analyzes PRs, runs static checks, opens issues, and coordinates multi-agent review flows.
  • Designed to run locally, in CI, or as a GitHub app.
• Why it matters:
  • It’s AI applied to DevEx and governance: quality gates, automated issue creation, platform detection, and multi-agent workflows instead of another “chat in your IDE” toy.

VeritasVault is a full surface: contracts, services, landing, docs, and IaC. vv-docs is the knowledge hub aggregating contract, service, and infra docs across the VV repos. :contentReference[oaicite:6]{index=6}

Repo: https://github.com/JustAGhosT/vv-docs :contentReference[oaicite:7]{index=7}

• Stack:
  • Docusaurus + MDX, Node, Yarn
  • Mermaid diagrams, Algolia DocSearch
• Role:
  • Aggregates docs from: vv-chain, vv-chain-services, vv, vv-frontend, vv-gamification, vv-dev-tools, cognitive-mesh, etc.
  • Includes auto-sync pipelines via GitHub Actions.
• Signal: I actually implement unified context: code, infra, and docs wired into one documentation hub.

Repo: https://github.com/JustAGhosT/vv-landing :contentReference[oaicite:8]{index=8}

• Stack:
  • Next.js + React + TypeScript
  • Tailwind CSS, CSS modules
  • CI/CD with tests, linting, type-checks and deploys to VeritasVault.net.
• Role:
  • The public marketing site for VeritasVault.ai.
  • Uses structured types and components to keep product messaging and docs aligned.

Some VV repos are more infra and platform-facing:

• vv-iac – Azure IaC for VeritasVault (Bicep/ARM, policy definitions, deployments).
• vv-* family referenced in vv-docs – chain services, gamification, frontend, dev-tools, etc, documented from a single hub. :contentReference[oaicite:9]{index=9}

These together show how I think about product as a system: contracts → services → docs → infra → governance.

Repo: https://github.com/JustAGhosT/PhoenixRooivalk :contentReference[oaicite:10]{index=10}

PhoenixRooivalk is a serious counter‑UAS defense monorepo, not just a slide deck.

• Monorepo & tooling:
  • Turborepo + pnpm workspace
  • Linting & quality: ESLint, Prettier, Husky, Clippy, cspell
  • CI: GitHub Actions, CodeQL, Netlify deploys
• Apps:
  • apps/docs — Docusaurus technical documentation
  • apps/marketing — Next.js 14 marketing site (threat simulator, ROI calculator, interactive demos)
  • apps/api — Rust (Axum) API server
  • apps/keeper — Rust chain keeper
  • apps/evidence-cli — Rust CLI for evidence flows
• Rust crates:
  • crates/evidence — core evidence logging
  • crates/anchor-solana / anchor-etherlink — dual-chain anchoring (Solana + Etherlink)
  • crates/address-validation — address safety checks
• Why it matters:
  • It shows I can take a hardware-adjacent, security-sensitive domain and turn it into a properly structured Rust+TS platform: docs, UI, API, blockchain layers, and dev tooling in one place.

Yes, Rust is very much part of my current workflow.

These aren’t theory projects. They sit on real land, real livestock, and real partners.

Repo: https://github.com/JustAGhosT/farm-business-plan :contentReference[oaicite:11]{index=11}

• Stack:
  • Next.js application (Node/TS)
  • Tailwind CSS, Jest tests
  • CI/CD with GitHub Actions, deployed to Netlify (farmplan.netlify.app)
• Role:
  • Agricultural business plan template & farm management tool.
  • Mixes structured docs, financial models, and a web app for ongoing planning.

Repo: https://github.com/JustAGhosT/zeeplan :contentReference[oaicite:12]{index=12}

• Stack:
  • TypeScript-heavy Next.js-style codebase (Next config + TS, ~70% TS, ~29% CSS).
• Content:
  • Full partnership proposal (Zeerust, SA): current state, 5‑year transformation, equity splits, ROI analysis, risk mitigation, version history.
• Why it matters:
  • Shows how I negotiate & design performance-based, zero-cash entry equity structures and document them like a system design.

Repo: https://github.com/JustAGhosT/pigpro :contentReference[oaicite:13]{index=13}

• Stack:
  • Full-stack TypeScript monorepo
  • Vite-based frontend, Vitest tests, Tailwind config, Node backend folder structure
• Role:
  • Domain-driven structure for pig production operations.
• Signal:
  • I treat farm ops as a proper software domain: typed modules, tests, shared packages, not a one-off spreadsheet.

Repo: https://github.com/JustAGhosT/goat-farming-guide :contentReference[oaicite:14]{index=14}

• Stack:
  • Backend: C# Azure Functions API using Azure Cosmos DB
  • Frontend: Next.js + React
  • Infra: infra/bicep for Azure resources; Contentful headless CMS integration
• Purpose:
  • Comprehensive goat farming knowledge site: health, feeding, economics, mistakes, investor info, DIY milking stand, etc.
• Why it matters:
  • Shows how I combine .NET cloud functions + TS frontend + headless CMS + Bicep to ship a content-heavy domain product.

Repo: https://github.com/JustAGhosT/crisis-unleashed-app :contentReference[oaicite:15]{index=15}

• Stack:
  • Frontend: frontend-next – React / Next.js + TypeScript + Tailwind
  • Backend: backend – Python FastAPI
  • Tooling: Docker (frontend/backend), pnpm workspace, Jest, Pyright, k6 perf tests
• Role:
  • Modular full‑stack template with auth, API, tests, CI, clean architecture.
  • Also doubles as a digital TCG / chain‑integrated concept to keep it interesting.

Repo: https://github.com/JustAGhosT/home-lab-setup

• Stack:
  • PowerShell automation
  • Azure networking + infra
• Scope:
  • P2S VPN, NAT, DNS, certificate lifecycle, monitoring/alerts, cost guardrails (as per repo summary).
• Why it matters:
  • Shows my defaults for secure, observable cloud networking and cost control in a small but realistic environment.

• Platform: cognitive-mesh
• Governance UI: CognitiveMeshUI
• DevEx / CI automation: autopr-engine

• Docs hub: vv-docs
• Public surface: vv-landing
• Infra & supporting repos: vv-iac, vv-chain, vv-chain-services, vv-gamification, vv-frontend, vv-dev-tools (all wired together via vv-docs). :contentReference[oaicite:16]{index=16}

• Regenerative farm partnership & ROI structure: zeeplan
• Crop-agnostic farm planning app: farm-business-plan
• Operational pig production system: pigpro
• Goat farming API + content platform: goat-farming-guide

• Full Rust + TS monorepo with docs, marketing, API & chain anchoring:
  PhoenixRooivalk

• Full-stack template (React + FastAPI, CI, tests, perf tooling):
  crisis-unleashed-app
• Infra defaults & Azure plumbing:
  home-lab-setup

If you only have 10 minutes, skim the READMEs in:

• cognitive-mesh
• PhoenixRooivalk
• vv-docs / vv-landing
• farm-business-plan
• zeeplan
• pigpro

Languages & runtimes

• TypeScript / JavaScript – Next.js, React, Node, monorepos (Turborepo, pnpm, Vite) :contentReference[oaicite:17]{index=17}
• Python – FastAPI backends, automation tools, CLI, PR automation (autopr-engine, crisis-unleashed-app) :contentReference[oaicite:18]{index=18}
• C# / .NET – cognitive-mesh, Azure Functions API (goat-farming-guide) :contentReference[oaicite:19]{index=19}
• Rust – API services, chain keepers, CLI & core crates in PhoenixRooivalk :contentReference[oaicite:20]{index=20}

Cloud / infra

• Azure (Functions, Cosmos DB, Static Web Apps, VNets/VPNs, DNS) :contentReference[oaicite:21]{index=21}
• Netlify, Vercel and static exports for marketing / doc sites :contentReference[oaicite:22]{index=22}
• IaC via Bicep / ARM (goat-farming-guide, vv-iac), plus homelab PowerShell.

Architecture & delivery

• Monorepos with pnpm / Turborepo where it makes sense
• Clean boundaries between frontend/backend/docs/infra
• CI/CD with GitHub Actions: tests, linting, type-checking, deploys on most public-facing repos :contentReference[oaicite:23]{index=23}
• Policy-as-code orientation (vv‑*, home-lab, vv-iac), and eval/telemetry hooks on AI systems.

• LinkedIn: linkedin.com/in/juriesmit
• Phoenix VC site: phoenixvc.tech
• GitHub: you’re here – issues/PRs are welcome on anything above.