Logo
Mev, Llc.

Senior Full-Stack Engineer (Backend Focus, Agentic AI)

Mev, Llc., Phoenix, Arizona, United States, 85003

Save Job

Project Overview

Our client is building an AI-driven learning platform that transforms technical documentation (PDFs, presentations, technical manuals) into interactive courses with automated video generation and intelligent AI tutoring. The platform serves pharmaceutical, manufacturing, and R&D sectors where subject matter experts need to convert complex technical content into training materials without relying on traditional learning and development teams.

We are looking for a Senior Full-Stack Engineer with strong backend expertise to be responsible for core backend services and agentic AI workflow integration. You'll architect and implement event-driven systems that orchestrate multi‑stage AI pipelines (document extraction, content structuring, video generation), integrate external AI services, and build reliable job processing with checkpointing and error recovery.

This role requires strong full‑stack development skills with a deep focus on the backend, hands‑on experience in building or integrating agent‑based workflows, and familiarity with LLM‑based systems.

Technical Details:

Database:

PostgreSQL (Supabase), Row‑Level Security (RLS)

AI/LLM:

OpenAI API, LangChain/LangGraph, Arize AI/Phoenix

Job Queues:

Redis/BullMQ or AWS SQS

Storage:

AWS S3 with signed URLs

Email:

Resend or SendGrid

Error Monitoring:

Sentry

Analytics:

Mixpanel, Amplitude, or PostHog

Infrastructure:

AWS (Lambda, ECS Fargate, CloudFront), Docker, Terraform/Ansible

CI/CD:

GitHub Actions

Observability:

Sentry, CloudWatch

Job Highlights:

Cutting‑edge agentic AI application combining multiple AI services into cohesive workflows

Real‑world impact, helping technical experts create accessible learning content

Opportunity to work with OpenAI, Anthropic, and specialized AI services (document extraction, video generation)

Greenfield project with modern tech stack and architectural freedom

Team Composition: Architect, Software Engineer, Designer, QA Engineer, DevOps Engineer, BA, Delivery Manager

Responsibilities

Build backend services and APIs that orchestrate multi‑step workflows with checkpointing and error recovery

Design and implement agentic systems for course generation, chatbot functionality, and content processing

Operationalize prompts and LLM workflows designed by client‑side AI and instructional design experts

Integrate third‑party APIs (LLM providers, video generation) with proper rate limiting, retry logic, and cost monitoring

Implement secure authentication flows, RBAC policies, and tenant isolation using Supabase Auth and RLS

Write database migrations, design efficient schemas, and optimize queries for performance

Contribute to frontend development when needed to support full‑stack features

Integrate application‑level monitoring (Sentry) and analytics tools (Mixpanel/Amplitude/PostHog) with backend services

Participate in architecture decisions, code reviews, and sprint planning

Collaborate with the solution architect on technical design decisions

Work with DevOps engineer on deployment requirements and observability tooling

Partner with a frontend‑focused engineer on API contracts, data models, and full‑stack feature integration

Qualifications

5+ years of full‑stack development experience with a strong backend focus (Node.js, TypeScript)

Experience using AI‑powered productivity tools (Cursor, VS Code with Copilot, or similar AI‑enhanced IDEs) and LLMs for research and problem‑solving (ChatGPT, Claude, etc.)

Experience building or integrating agentic workflows or LLM‑powered applications

Hands‑on experience with LangChain, LangGraph, or similar orchestration frameworks

Familiarity with OpenAI and/or Anthropic APIs, prompt engineering, and structured output validation

Working knowledge of React and Next.js (able to contribute to full‑stack features when needed)

Solid understanding of PostgreSQL (schema design, migrations, query optimization)

Experience with event‑driven architectures, job queues (Redis/BullMQ, AWS SQS), and asynchronous processing

Familiarity with NestJS or similar backend frameworks (Express, Fastify)

Experience with authentication, authorization, and multi‑tenant data isolation patterns

Strong problem‑solving skills and ability to work autonomously with minimal oversight

Clear written and spoken English for async collaboration and documentation

Nice to Have

Experience with Supabase (Auth, RLS, Realtime)

Familiarity with LLM observability tools (Arize AI, Phoenix, LangSmith)

Experience with document processing pipelines or OCR integration

Knowledge of video generation APIs (Synthesia or similar)

Familiarity with error monitoring (Sentry) and product analytics tools (Mixpanel, Amplitude, PostHog)

Understanding of prompt engineering principles and best practices

Experience with evaluation frameworks for LLM‑generated content

Background in EdTech, content generation, or learning platforms

Familiarity with Tailwind CSS and modern UI development patterns

GitHub‑based workflows and conventional commits

Work & Life Balance 20 paid working days off

Soulful corporate events and team buildings

Opportunity to work remotely

All the necessary equipment

A comfortable and modern office

Accounting services

Health insurance

$600 extra for the health care, sports or mental health

10 working days of sick leave

Professional Development $600/year you can spend on:

Improvement of qualification/ certification

[ our values/ ] What Keeps Us Sharp and Steady We cut the fluff and keep it real—owning our work, refining our craft, and always leveling up. No excuses, just a team that’s in it together, pushing boundaries and delivering results.

Pragmatic

Theory’s cute, but results pay the bills. Perfect’s a fairy tale—done right is what counts.

Adult

No drama, no sugarcoating. We own our wins and our screw‑ups, and treat failure like free tuition.

Craftspeople

We’re not assembly‑line coders. We build, we refine, and we don’t put a fresh coat of paint on bad code.

Falling behind? We leave that to others. Leveling up fast, chasing down answers, and never settling for “I don’t know.”

Great Allies

No lone wolves here. We listen before we talk, communicate like grown‑ups, and make the ride better for everyone.

Headquartered in the USA and Ukraine, but our team? We're everywhere. We’re fully remote, so you can work wherever you’re most productive.

#J-18808-Ljbffr