Full-Stack Developer

What you bring

• 4+ years building and running production web applications, across both back end and front end.
• Strong Python: FastAPI or a similar framework, Pydantic, async code, and pytest.
• Hands-on MongoDB: schema design, indexes, aggregation, and idempotent writes under retries and concurrency.
• Strong React and TypeScript: routing, data fetching and caching, forms, and Tailwind or a component library.
• Experience with third-party APIs and webhooks: OAuth, signature checks, retries, rate limits, and partial failures.
• Stripe experience: subscriptions, Checkout, webhooks, and ideally usage-based billing.
• Comfort deploying and debugging containers in the cloud (Cloud Run, Docker, Vercel), and reading traces and logs to find a root cause.
• Good judgment with personal data and email: privacy rules, consent, and why one wrong send matters.
• Clear written communication: plans on issues, PRs that explain what changed and why, and honest status.

Your first 90 days
By day 90 you own one area of each product and can ship to either without help. 1. Days 1–30: Run Zipisa and WebHawk locally with their full test suites passing. Ship 3–5 small fixes to production in both. Learn the send guardrails, the task drains and the billing flows. 2. Days 31–60: Own a mid-sized feature end to end, such as a new CRM integration or a dashboard screen, from Linear plan to production. Help debug and fix production issues as they come up. 3. Days 61–90: Lead a cross-product piece of work, such as WebHawk visitor data inside Zipisa or metered billing. Leave a runbook or regression test behind for each incident you handle.

What you'll do

• Build features end to end. Take a Linear issue from plan to production: FastAPI routes, Mongo models and indexes, React screens, tests, and a clear PR.
• Keep the email pipelines safe and deliverable. Work on Zipisa's mail queue and WebHawk's KumoMTA sending: idempotent jobs, send guardrails, warm-up ramps, unsubscribe handling, bounce and reply processing.
• Ship and maintain integrations. Build and fix CRM and calendar connections (Follow Up Boss, kvCORE, HubSpot, Pipedrive, Salesforce, Calendly, Zapier, webhooks) and the data providers behind lead sourcing and validation.
• Own billing flows. Implement Stripe subscriptions for Zipisa and metered, pay-per-use charging for WebHawk, with a usage ledger that is correct under retries.
• Build the dashboards. Visitors, inbox, campaigns, domains, integrations and settings screens in React and TypeScript, wired to the API with loading, empty and error states handled.
• Work on the AI layer. Wire LLM features (reply drafting, lead decisions, the in-app assistant) into the product, with evals that catch regressions before customers do.
• Run what you ship. Deploy to Cloud Run and Vercel, watch Logfire traces, run scheduled task drains, debug production issues, and leave behind a regression test or runbook for every incident.
• Protect personal data. Handle PII, consent signals and do-not-contact rules carefully, and keep secrets and raw payloads out of logs.

Nice to have

• Email deliverability: SPF, DKIM, DMARC, mailbox warm-up, MTAs such as KumoMTA or Postfix, and reading bounce and complaint signals.
• LLM product work: prompt design, tool-using agents, evals, and controlling cost per conversation.
• Real estate tech: Follow Up Boss, kvCORE/BoldTrail, MLS data, or how agents and brokerages work leads.
• Ad-tech or identity resolution: pixels, first-party data, consent signals such as GPC, and match rates.
• TanStack Start or Next.js, server-side rendering, and technical or programmatic SEO.
• Infrastructure: Cloudflare (Workers, R2, cron), Redis, queues, Terraform, or running a Linux box.
• Experience as an early engineer at a startup, owning a product with little hand-holding.
• Daily use of AI coding agents such as Claude Code, with the habit of reviewing their work skeptically.

About the role
We're hiring a full stack developer to build and run two products that share one stack: Zipisa, an AI assistant that finds sellers and books appointments for real estate agents, and WebHawk, which identifies website visitors and turns them into leads. You'll ship across Python/FastAPI services, MongoDB, and React/TypeScript dashboards, and own features end to end, from the API through the email pipeline to the screen the customer sees. Both products send real email to real people and handle personal data and payments, so we care about safe defaults, tests, and observability as much as speed.

Zipisa- An AI inside sales agent (ISA). It sources seller-intent leads by ZIP code, emails homeowners, replies and follows up, and books appointments on the agent's calendar. WebHawk

Webhawk - Website visitor identification. A pixel on the customer's site resolves anonymous visitors into people, then delivers them to the customer's CRM and sends behavioral follow-up email.