Amy Villa — AI Automation & Technical Solutions Portfolio

AI Automation & Technical Solutions Engineer
API-driven automation · Developer-facing systems · Product engineering

GitHub profile · Email

I build API-driven automation and developer-facing systems that make complex workflows easier to implement, debug, and operate.

This repository is the evidence interface around a set of public engineering studies. It is designed for three review depths: a fast recruiter scan, a deeper hiring-manager walkthrough, and source-level technical review through GitHub, tests, CI, and code paths.

Review paths

Path	Time	What to inspect
Recruiter	~2 min	Specialization, strongest systems, core capabilities, contact
Hiring manager	5–10 min	Architecture, tradeoffs, failure states, testing, implementation boundaries
Technical reviewer	As needed	Public repositories, code paths, tests, CI, and reproducible local demos

Start here

System	What it demonstrates	Public evidence
MailTrace DX Lab	Developer observability · webhook reliability · API debugging	HMAC, replay resistance, idempotency, retries, event ordering, raw payloads, SQLite, strict TypeScript, Playwright
ForgeFlow AI Automation	AI automation systems	Structured execution, optional AI provider, deterministic fallback, persisted state, human review, FastAPI
Signal Engine — this repository	Technical narrative · evidence interface	React/TypeScript, tested interaction state, recruiter path, product/system explanation
ClearRoute API	API design · workflow state	Pydantic validation, explicit transitions, role rules, audit events, predictable error contracts
AccessPath Console	Product engineering · accessibility	Keyboard UX, semantic HTML, validation recovery, live feedback, axe/Vitest checks

Developer Experience & Agent Tools

Five public Python/FastAPI projects extend the portfolio's engineering evidence:

Project	Review focus	Source
DX Orbit	Explainable OpenAPI scorecards, revision comparison, and report export	Repository
ToolTrust	JSON Schema argument validation, permissions, and confirmation traces	Repository
DevStart	API onboarding scenarios and explicit recovery guidance	Repository
HookForge	Duplicate delivery, stale-event rejection, and webhook failure simulation	Repository
SignalDesk	Actionable documentation feedback and issue draft export	Repository

The website includes source-backed case studies at /projects/<project-name>. The Python applications run locally; the site does not claim to host their APIs. Each case study documents its limitations and links to tests and CI results.

Portfolio architecture

Browser
  │
  ▼
React + TypeScript evidence interface
  ├── Home / specialization + technical work
  ├── Recruiter Fast Path
  ├── Interactive project studies
  └── Signal Lab
        │
        └── public engineering repositories
             ├── MailTrace: observability + reliability
             ├── ForgeFlow: AI automation systems
             ├── ClearRoute: API contracts + workflow state
             └── AccessPath: accessibility + recovery

Main code map

Path	Responsibility
client/src/App.tsx	Application routing and resilience boundary
client/src/lib/productEvidence.ts	Typed candidate and project evidence used across recruiter-facing surfaces
client/src/lib/signalEngine.ts	Pure, tested scene/progress navigation math
client/src/pages/Home.tsx	AI Automation & Technical Solutions hero, project catalog, and evidence actions
client/src/pages/RecruiterProof.tsx	Fast review path for engineering evidence and implementation conversations
client/src/pages/DemoPage.tsx	Interactive project-study surfaces
client/src/pages/SignalLab.tsx	Stateful workflow simulation
scripts/verify-product-engineering.mjs	Recruiter-signal, privacy, portability, and repository-hygiene checks
scripts/verify-recruiter-proof.mjs	Shared evidence and review-path checks
scripts/verify-build-budget.mjs	Production JavaScript/CSS bundle budgets
.github/workflows/verify.yml	Tests → hygiene → typecheck → recruiter proof → build → bundle budgets

Key product decisions

• Specialization is immediate. The first scene communicates AI Automation & Technical Solutions instead of making a reviewer infer the positioning after several scroll states.
• Evidence is source-first. Strong claims point to public repositories, tests, CI, or documented implementation boundaries.
• Failure behavior is part of the story. The linked systems expose degraded state, validation errors, retry behavior, or recovery paths instead of presenting only success cases.
• Human judgment remains visible. Automation is presented as a system with review points, not as unsupported autonomy.
• Different repositories prove different skills. Observability, AI automation, API design, and accessibility are intentionally separated instead of repeated as variations of the same tutorial.

Verification

pnpm install --frozen-lockfile
pnpm test:portfolio
pnpm verify:product-engineering
pnpm check
pnpm verify:recruiter-proof
pnpm build
pnpm verify:bundle

The pull-request workflow runs the same critical path. The production bundle guard caps total JavaScript at 300 KiB raw / 95 KiB gzip and CSS at 180 KiB raw / 45 KiB gzip so performance regressions become visible during review.

Deployment status

The current source and generated Vercel-ready shell are maintained in this repository. The existing public Vercel domain was previously deployed through a manual shell and is not treated as current execution evidence until it is refreshed from this source.

That distinction is intentional: a deployed URL is useful evidence only when the running version can be tied back to the reviewed source and verification path.

Run locally

pnpm install --frozen-lockfile
pnpm dev

Candidate context

AI Automation & Technical Solutions Engineer
Native English + Spanish · Remote collaboration

My paid freelance experience includes digital systems, AI-assisted workflows, process organization, bilingual delivery, and remote operational handoffs. The software repositories here are self-directed public engineering work; I do not present that history as years of formal full-stack employment.

Evidence boundary

These projects use sample data and demonstrate implementation choices and engineering judgment. They do not claim production users, enterprise scale, customer outcomes, formal security certification, accessibility compliance, or production experience beyond what the repositories explicitly document.

Contact

Amy Villa · amyv.dev@gmail.com · GitHub