Dan Raymond
I take hard problems from architecture through reliable, secure production.
First lead of a new Integration function.
I led the team responsible for the boundary between application development and infrastructure: defining integration requirements, coordinating releases, mentoring a multidisciplinary team, and staying hands-on while architecting the platform that turned a brittle multi-week Suite build into a repeatable automated capability.
System & Platform Architecture
Own architecture across APIs, state, data boundaries, runtimes, deployment topology, failure modes, and the operational contracts that keep complex systems maintainable.
Agentic & AI Systems
Design bounded autonomy, durable work, retrieval, model routing, evaluation, tool boundaries, verification, and recovery as production systems rather than prompt-layer demos.
Applied AI & Automation
Turn ambiguous operational problems into measurable systems: document intelligence, ML classification, browser automation, workflow orchestration, and human review gates.
Reliability, Security & Technical Leadership
Design for failure, security boundaries, reproducibility, observability, and long-term ownership while aligning engineers, product, operations, cyber, and leadership around the architecture.
The architecture should be visible before the paragraph.
Six systems selected for professional impact, operational depth, and inspectable proof. Employer work stays sanitized; independent and public systems show real interfaces wherever possible.
ATA AI Platform
Production AI architecture across document intelligence, ML mapping, QA automation, reproducible runtimes, and adoption.
Della
Persistent autonomous AI with explicit authority, durable execution, evidence-backed verification, restart recovery, and bounded browser acquisition.
BuildingRelay
Deployed AI case-management platform with dual-path reasoning, approvals, building memory, and safety-first external-action controls.
OpenDuck
Cloud-hosted agent operations platform with LangGraph orchestration, campaign experimentation, explicit approval boundaries, and private model routing.
JobFinder
Evidence-grounded application automation with deterministic ATS adapters, rerender verification, recovery recipes, and guarded browser-agent escalation.
WebbDuck
Public multi-backend local creative platform spanning SDXL, FLUX, Krea 2, and Qwen with capability-driven routing and identity workflows.
Architect for what comes after the demo.
Retrieval, orchestration, evaluation, infrastructure, security, approvals — the work that lets teams actually depend on the system.
Autonomy with kill switches.
Governance, auditability, and human approval where they matter — useful agents in real organizations, not lab toys.
Measurable behavior over vibes.
Regression checks, eval harnesses, observable workflows. If you can't measure it, you don't really ship it.
Adoption is half the work.
The best platform is the one people understand, trust, and actually use. Onboarding and enablement aren't afterthoughts.
Production tools across hosted and self-hosted systems.
Model APIs and local runtimes where each fits, backed by infrastructure and orchestration that can be observed, secured, and maintained.
Need ownership for a hard AI or platform problem?
I'm open to Principal / Staff AI and platform roles, architecture work, and senior consulting engagements where hands-on technical ownership matters.