Full-Stack Engineer — AI Training Simulator for Posh
About the Company:
Posh AI is an AI-powered training platform that uses realistic voice and chat simulations to help employees practice real-world customer and workplace scenarios. The platform provides instant AI feedback, performance scoring, and analytics to improve skills, confidence, and readiness.
About The Role:
We’re looking for a full-stack engineer to join the team building Posh AI Simulator, a B2B platform that lets learners practice real-world conversations with AI-powered avatars via text chat and live voice. The platform serves financial services companies and L&D teams who need scalable, measurable role-play training — onboarding, compliance practice, and coaching at scale.
You’ll work across the entire stack: a React SPA, a Supabase Postgres backend with 50+ Deno edge functions, and deep integrations with large language models, Auth0, and enterprise LMS standards.
What You’ll Work On
- Building and iterating on learner-facing simulation experiences (chat and real-time voice) and the admin/content-creation studio
- Designing and maintaining Postgres schemas, RLS policies, and RPCs for a multi-tenant data model
- Writing Deno edge functions that orchestrate LLM calls — chat, voice sessions, evaluation, document parsing, insight generation
- Integrating with enterprise learning systems via xAPI and LTI 1.3
- Shipping analytics and insights features that help team leads understand learner performance
- Maintaining a clean, type-safe TypeScript codebase with strong test coverage using Vitest and React Testing Library
Tech Stack
- Frontend: React 18, TypeScript, Vite, React Query, Zustand, Tailwind CSS, shadcn/ui
- Backend: Supabase (Postgres, Auth, Storage, Edge Functions), Deno, Row-Level Security
- AI: LLMs for chat, real-time voice (WebRTC), embeddings, and tool calling
- Auth: Supabase Auth, Auth0 enterprise SSO
- LMS: xAPI, LTI 1.3, HMAC-signed launch URLs
- Infra: AWS S3 (static hosting), GitLab CI, Groundcover observability
- Email: Resend with React Email templates
What We’re Looking For
- Strong TypeScript skills and experience with modern React patterns (hooks, React Query, Zod)
- Comfort working with Postgres — schema design, migrations, RLS, and stored procedures
- Experience building or integrating with LLM APIs
- Ability to write edge functions / serverless handlers (Deno or Node background is fine)
- Product instinct — you can translate a vague requirement into a well-scoped implementation
- A bias toward simple, well-tested code; you write tests as part of the work, not after
- Advanced English proficiency (B2–C1 or higher)
Bonus Points
- Experience with real-time audio/WebRTC or voice AI
- Familiarity with xAPI, SCORM, or LTI integrations
- Multi-tenant SaaS background
- Financial services or regulated-industry domain knowledge
What to Expect
You’ll ship features that directly affect how thousands of learners practice and improve. The codebase is modern and well-structured, deployment is automated, and the team moves fast. You’ll have real ownership over both product direction and technical decisions.
Nice to Have
- Deno experience
- Supabase + RLS
- WebRTC
- xAPI/LTI/SCORM
What We Offer
- We welcome new ideas and allow you to make an immediate impact on the team.
- Flexible Paid time off (PTO for any reason, including sick days (no specified limits) and flexible work schedule.
- Health/Sport Budget.
- Fully remote.
As this is an English-speaking role, we kindly ask that all applications be submitted with a CV in English.
