Alysson Alcântara

AI-native role expanding the scope of traditional dev: discovery with stakeholders, product definition, and delivery via AI-driven development.

• Built the internal franchise management portal for 208 franchisees in partnership with stakeholders in <4 weeks, using Claude Code connected to internal MCPs we built for Seazone services; shipped full cleaning module, franchisee perspective mode, guided onboarding, and feature flags via GrowthBook
• Built a market intelligence platform covering 60+ Brazilian cities with real data from Airbnb and Booking, integrating AWS Athena Data Lake via MCP and a conversational agent (Vercel AI SDK + Gemini 2.0 Flash); selected as the technical foundation for Wallet Helm, Seazone's market intelligence product for short-term rentals
• Co-built an AI-driven development workflow that turns Discovery handoffs into ready-to-code workspaces, adopted as the standard across squads and SRE; also contributed to the internal AI-native platform that replaced Outline and Jira company-wide

seazone.com.br, Seazone's public vacation rental booking site, mobile-first. Frontend Next.js + TypeScript + Tailwind, backend Python.

• Implemented 3DS authentication on the on-site checkout frontend (shifts chargeback liability to the card issuer on confirmed transactions) as part of the team's migration from Asaas to Tuna (Brazilian payment orchestrator with Cybersource anti-fraud) — triggered by Asaas's 50% rejection rate in production and a R$50k loss from a coordinated chargeback wave; also integrated PayPal and Pagarme. Monthly chargeback volume after migration: R$140k → <R$10k (~93%)
• Authored 16 of 41 experiments in the project (including experiments on search filters and new homepage running at scale) as owner of the experimentation infrastructure: TestManager service with multi-provider architecture (PostHog for experiments, GrowthBook for feature flags), server + client with experiment caching and server-side flag bootstrap — architecture later adopted as the standard across other Seazone projects
• Delivered +13% in click-to-pay and +9% in paid reservations over 8.8k users (PostHog, Mar–Apr 2025) as owner of the "without-login checkout" iterated through 3 sequential experiments; also owned the late-checkout flow (BE + FE)
• Cross-stack: contributed to the Python backend with OpenSearch for queries on the city filter in promotions, WhatsApp integration for abandoned cart, and server-side env-gating of PostHog
• Technical SEO: implemented JSON-LD/Schema.org, dynamic sitemap, and structured per-route meta tags on the public booking site — ~118k organic impressions and ~8.7k clicks/month via Google (Search Console, last 3 months)
• Performance and infra: optimized build pipeline (bun early adopter → pnpm, multi-stage Dockerfile, deps cleanup) — image 6.7GB → 388MB (~94%), build ~50% faster; sequential Next.js upgrades (13, 14, 15)

Intensive frontend engineering bootcamp. Led 3 consecutive cohorts of ~40 students (ages 20 to 60+) in 3-month cycles, totaling ~120 students during the period.

• Placed 5 of my students in international roles in Germany before course completion (~10 total employed before finishing), in a bootcamp with 98% reported employment rate within 6 months during the period
• Managed the full lifecycle of each cohort: technical lectures, 1:1 individual mentorship, ongoing evaluation, assignment grading, personalized support for at-risk students, and judging of final multi-group projects
• Taught advanced content in frontend performance (memoization, code splitting), state management (Redux, Context API), TypeScript, and Next.js rendering strategies (SSR/SSG/ISR)

Seazone's internal multi-role platform that digitalized operations previously run on spreadsheets and WhatsApp. TypeScript + Python monolith built by a lean team; I joined as one of the first frontends.

• R$1M+ disbursed to 159 partners since launch as sole developer end-to-end of the Partners portal: PWA-first dashboard for Seazone's referral program (partners earn a percentage on reservations they referred), replacing the manual report previously sent via Google Scripts + spreadsheets
• Sole developer of the pre-checkin module end-to-end: original feature, API integration with cross-stack contributions to the Python backend, and three subsequent refactors; digitalized a pre-arrival process previously run on spreadsheets + WhatsApp
• Bootstrapped the Jest testing infrastructure and acted as primary frontend reviewer on the team, performing more code reviews than PRs I authored during the period

B2B SaaS for gamified employee performance reviews in retail; spinoff of a Hacking Rio 2019 winning project, 4-person founding team.

• Led end-to-end design and brand identity: UX research, wireframes, design system, and go-to-market materials
• Served as account manager for the first B2B paying customer (food retail), with a monthly recurring contract delivering design and weekly analytics
• Designed and shipped a gamified ranking system for in-station feedback, used during the team's 3-month validation pilot across 3 BR Distribuidora gas stations, with a nationwide scale thesis for the gas station network

UENF junior enterprise for external clients (CNPJ, ~R$8k cash flow, collective governance).

• Delivered ~3 projects for external clients in React (UX/UI design + dev) under senior-student mentorship, with hands-on grounding in Git, code review, and exposure to Java and C; first professional experience before Avaliei