portfolio.init --prepared-for Jason Deegan @ Gain Clarity AI

Applied AI Builder
focused on agentic workflows
and practical AI products

I build AI-assisted internal tools, workflow automations, dashboards, and agentic systems — the same practical software bottleneck space Gain Clarity AI is attacking.

internal_tools workflow_automation agentic_systems owned_deployments
why_this_role

Why this role fits

"Applied AI is where I naturally spend my time."

Internal tools mindset

Turning LLMs, APIs, and automation into workflow software people can actually use.

Deployment-aware builder

Thinking beyond demos: ownership, security boundaries, verification, and rollout friction.

Fast iteration rhythm

Explore, build, verify, document gaps, and polish — without pretending prototypes are production.

blixnex_product

Blixnex: AI product work for local service businesses

"Applied AI for local service businesses: faster responses, more leads, smarter operations. Some features are demos or prototypes."
blixnex.com
Blixnex homepage preview
Hover to explore. Open the full Blixnex site below.

Customer communication

Chat, SMS, and voice receptionist concepts.

Workflow automation

Booking, follow-up, and lead handling.

Demo render generation

High-fidelity local-service previews as a product feature.

hermes_orchestration

Hermes orchestration workflow

"The hero shows the live terminal feel. This map shows the system: objective → orchestration → execution → verification → review."
Bill / voice promptSets objective, constraints, approval bar
capture + route
TelegramFast command surface
TypelessVoice capture when useful
Hermes Agent on VPSContext, memory, skills, tools
delegate + inspect
Claude Code / CodexImplementation agents
Repo + toolsFiles, CLI, browser, deployment
Human checkpointsBill approves important calls
verify + report
TestsRun checks before claiming done
Screenshots / R2 artifactsVisual evidence and handoff assets
Linear updatesProject tracking and next steps
Simple English reportWhat changed, passed, remains

// how_i_use_it

Hermes is my Telegram-connected orchestration agent running on my VPS. It coordinates Claude Code and Codex, checks outputs, verifies tests and screenshots, updates project tracking, and reports back in simple English.

// supporting_infrastructure

Hardened VPS environment
VS Code SSH workflow
Cloudflare Tunnel access
Agent orchestration
Restricted permissions
Human-in-the-loop review

// Framing note: this is my personal applied AI development workflow, not a claim of enterprise production infrastructure.

demo_render

Blixnex demo render: premium HVAC concept

"I built a system that generates high-fidelity demos for Blixnex. This render is a feature, not a separate product."
Blixnex demo render
Hover to explore. Open the full render below.

What it is

A generated website preview for an HVAC concept, built as a Blixnex feature.

What it proves

Business direction to polished visual demo with screenshots and review notes.

Honest status

Demo artifact, not a live customer site. Some imagery is placeholder.

proof_of_shipping

Proof of Shipping

"Evidence of progress matters most: working demos, tests, screenshots, and iteration loops. I focus on working artifacts, review links, tests, screenshots, and documented iteration instead of inflated metrics."
agentic_iteration

Idea to verified artifact

Broad objectives into code edits, browser checks, screenshots, and plain-English reports.

render_milestones

Blixnex previews as proof

Visual product direction with honest labels on prototype vs. live.

verification_loop

Tests, links, screenshots

Check links, capture evidence, document gaps before calling work ready.

project_rhythm

Linear + human review

Tracked updates, handoff notes, and human-in-the-loop review.

github_artifact

One supporting artifact

GitHub is one supporting artifact. The stronger signal is the workflow behind the work: scoped tasks, implementation, verification, and review.

clarity_fit

What I bring to Gain Clarity AI

Internal-tool mindset

I like software that matches how work actually runs — workflows, dashboards, automations, portals, and agents.

Speed + verification

Fast, but always checking outputs, links, screenshots, deployments, and claims before calling work done.

LLMs + APIs + operations

Connecting language models, APIs, backend automation, and human review into useful systems.

Plain-English handoff

What changed, what passed, what failed, what is uncertain, and what should happen next.

continuing_to_improve

Production engineering depth

Continuing to grow in production deployment patterns, backend reliability, security-aware internal tooling, cloud architecture, and production AI systems.

"Turning AI demos into daily workflows."

I'm looking for a role where I can help build practical AI products, learn quickly, and take ownership of real outcomes.