Full Stack Developer

Practis gets field sales teams ready before they meet a paying customer. The people who sell for pest control, lawn care, HVAC, home improvement, and insurance companies are often canvassers, technicians, and comfort specialists, not people with "sales" in their title. Most get a few days of shadowing and a script, then start working leads the company paid for.

Practis changes that order. New hires practice real conversations in private with an AI coach. They certify against the PRACTIS™ Method, our published standard for field sales conversations, and only then take live leads. Operators get evidence on the metrics they already track. We deliver it as a managed service: the customer provides access and context, and we build the content, run the launch, and report the outcome.

Practis runs on two Type Script codebases: a NestJS API and a single Expo app that ships to iOS, Android, and the web. New hires use it to hold spoken roleplay conversations with an AI customer, get coached, and certify. Managers use it to see who is ready, and operators use it to see what that readiness paid.

You will ship features across the whole stack, from database schema to the screen a technician opens in a truck between appointments. The work spans real-time voice, AI coaching, multi-tenant data, and mobile releases, and you will work alongside AI coding agents as a normal part of the job.

• Build product features end to end: API endpoints, database schema and migrations, and the mobile and web screens that use them.
• Work on real-time voice practice built on Live Kit, and on the AI coaching and scoring behind it.
• Keep each customer's data isolated, using PostgreSQL row-level security and careful API design.
• Ship to the App Store, Google Play, and the web, including over-the-air updates.
• Write tests, review pull requests, and keep CI green.
• Instrument features with Post Hog and Open Telemetry, so we know what customers use and where it breaks.
• Use AI coding agents well: write clear specs, review their output critically, and keep the quality bar high.
• Help shape Linear tickets and their acceptance criteria before they are built, not just after.

• 4 or more years of professional software engineering, shipping production web or mobile applications.
• Strong Type Script on both the front end and the back end.
• Production experience with React or React Native. Expo experience is a plus.
• Experience building APIs in Node.js, ideally with NestJS, and designing relational schemas in PostgreSQL.
• Comfort with Git-based workflows, pull request review, and CI/CD.
• Clear written communication: you can explain a trade-off in a pull request description or a ticket.

• Real-time audio, WebRTC, or Live Kit.
• LLM-backed features, prompt design, or model evaluation.
• Temporal, BullMQ, or other job and workflow systems.
• AWS infrastructure as code with CDK.
• An app you shipped through App Store review.

• API:
  NestJS 11, Type Script, Drizzle ORM, PostgreSQL with row-level security
• App:
  Expo SDK 54, React Native 0.81, expo-router, Tan Stack Query, Zustand, Zod, Native Wind
• Voice and real time:
  Live Kit, Twilio, Socket.
  
  IO
• Jobs and workflows:
  BullMQ on Redis, Temporal
• Auth:
  WorkOS
• AI:
  Anthropic models through the Vercel AI SDK
• Infrastructure: AWS (CDK, ECS, S3, Amplify), Git Hub Actions
• Observability:
  Post Hog, Open Telemetry, Firebase Crashlytics, New Relic

• Features you built running in production on both the API and the app.
• One feature owned from ticket to release, including tests and analytics.
• The full stack running on your machine, and one improvement to how we do that.

Kansas City metro preferred. Remote available for the right candidate in the US.