Founder, NULLWORKS | Operational Intelligence Systems Architect | Forward-Deployed Applied-AI Product Builder
Portfolio · OI Field Notes · Phoenix, Arizona
I help individuals and organizations turn disconnected AI tools, agents, expert knowledge, records, and workflows into a human-readable Operational Intelligence operating system.
The goal is not another chatbot. It is visible ownership, scoped workrooms, source-linked decisions, authority boundaries, review gates, continuity, telemetry, and final human control.
The AI engineer builds the worker. The OI architect builds the company.
Through NULLWORKS, I operate as a one-person forward-deployed discovery, orchestration, and prototyping unit. I direct AI as a digital workforce across operator discovery, workflow mapping, product framing, UX, architecture, implementation, debugging, deployment, source preservation, and telemetry.
The goal is not to pretend a rapid prototype is production-ready. The goal is to reduce uncertainty early and give specialist teams a validated working frame they can harden, secure, scale, polish, and operate.
An OISA designs the human-readable operating layer connecting:
- Human experts and final authority
- AI workers, models, tools, and automations
- Intake, routing, ownership, and work state
- Project workrooms and scoped context
- Source evidence and uncertainty
- Permissions, review gates, and escalation
- Continuity, decisions, corrections, and failures
- Telemetry, recovered capacity, and continuous improvement
AI is the tool. Operational Intelligence is the connected operating system. Human authority remains final.
DISCOVER → STRUCTURE → BUILD → VERIFY → IMPROVE
- Operator and customer discovery
- Workflow mapping and requirements
- AI/tool and authority inventories
- Product, UX, and dashboard framing
- Data and system architecture
- Functional prototype development
- Agent and workflow orchestration
- Deployment, debugging, and telemetry
- A validated real-world workflow
- Functional deployed software
- Explicit requirements and authority boundaries
- Structured operational knowledge
- Source-linked decisions
- Failure receipts and known unknowns
- A prioritized security, hardening, and scaling backlog
- A measurable before-and-after operating baseline
| System | What it proves | Live |
|---|---|---|
| LenderFlow / LENA | Direct broker discovery converted into a human-reviewed lender-fit and workflow beta with appetite, exceptions, routing, freshness, source receipts, and missing-information checks | Open beta |
| LegalFlow LF2 / KONRAN | Source-linked evidence workflows, searchable derivatives, timeline reconstruction, and human-expert authority | Open dashboard |
| PAPERGOBLIN | OCR intake, editable correction, validation, persistence, and reusable human-correction telemetry | Open prototype |
| ANVIL OI Production Studio / CUTSYNC | Structured creative intake, media handling, versioning, operator packets, licensing boundaries, and human review | Open beta |
| CHECKMATE | Recovery-first receipt OCR, editable reconstruction, and collaborative payment workflow UX | Open prototype |
- Several hours of direct discovery with a working broker
- Core workflow prototype in roughly three days
- Broader beta expanded over approximately two weeks
- Human review preserved; no automated lending or credit decisions
- Functional prototype built and deployed in approximately six hours during an airline flight
- Aging laptop, mobile debugging, limited workspace, and airplane Wi-Fi
- OCR failures and user corrections converted into reusable telemetry
- ET-10 electronics technician in high-throughput automation
- Mechanical, electrical, sensor, controls, OCR, networking, and conveyor fault isolation
- Traced a recurring 48-feed-chute failure to installation misalignment rather than software configuration
- Developed an independent human-centered AI logistics exception concept that reached senior USPS technology attention and was routed for technical evaluation
TAC OPS is an independent prototype concept, not an authorized or deployed USPS system. Views are my own.
- Let the Expert Expert
- When Your AI Assistant Becomes a Company
- OI SUITe: The Operating System Around the AI
- The AI Engineer Builds the Worker. The OI Architect Builds the Company.
- Horse Cart to Toyota: What Workflow Compression Is Worth
My software work is grounded in live operational systems where uptime, ambiguity, fault isolation, recovery, records, and human judgment matter.
- Industrial automation and logistics
- Applied AI and multi-agent orchestration
- OCR and document workflows
- Root-cause analysis
- Human-in-the-loop product design
- Source traceability and operational telemetry
- Full-stack Next.js / React / TypeScript / Supabase / Vercel delivery
- Forward-deployed discovery and implementation
- Operational Intelligence Systems Architect
- Forward-Deployed Engineer
- AI Solutions / Implementation Engineer
- Applied AI Product Engineer
- AI Platform Enablement Engineer
- Workflow Systems Engineer
- Human-AI Operations Architect
- OI discovery, blueprint, pilot, installation, and managed operations
Compress the mess.
Make the process observable.
Preserve the source.
Keep the expert in authority.
Measure the mess before claiming the compression.