Junior FDE (AI & Backend)

Context

We are building AI-native products on the Anthropic API and MCP, backed by a Django/DRF + PostgreSQL platform and a React/Next.js frontend, deployed on Azure. The team works in an AI-assisted workflow (agentic AI coding tools are a daily tool), so the Junior FDE must be able to review, understand and own every line that ships. This is a build-and-learn role: you work from tickets a senior has scoped, with code review on every change, and grow into owning full features.

What the role owns

Own scoped tickets end-to-end: build DRF endpoints and models, implement individual LLM workflows (prompts, tool calls, MCP tools), fix bugs from logs and traces, and test what you ship — then take feedback from review and cut repeat mistakes.

The actual workload (by share of time);

Backend features & APIs (35%)

DRF models, serializers, viewsets, migrations on PostgreSQL

• e.g. Add a filter + pagination to the cases list API

• e.g. New field with migration and serializer

LLM workflows, skills & MCP connectors (25%)

Prompts, tool use, agent skills, custom MCP tools/connectors

• e.g. Add an MCP tool that fetches a matter's documents

• e.g. Write a skill for the intake checklist

• e.g. Tighten the extraction prompt

Async jobs & integrations (15%)

Celery/Redis tasks, third-party APIs, auth (JWT/OAuth2)

• e.g. Move document summarisation to a Celery task with retry

Bugs & debugging (15%)

• e.g. Task stuck in pending — find why

• e.g. 500 on empty payload

Tests, review & light frontend (10%)

pytest, PR cycles, small React/Next.js fixes

• e.g. Add tests for the new endpoint

• e.g. Wire the new API into the Next.js list page

Priority reality: tickets arrive with mixed urgency. A Junior FDE flags blockers early, asks before guessing, and escalates anything touching auth, data or production to a senior rather than fixing it alone.

Tech stack

• AI/LLM engineering: Anthropic API, MCP (Model Context Protocol), agentic AI coding tools (CLI/IDE), reusable agent skills (SKILL.md-style), custom agents/subagents, custom MCP connector creation, prompt engineering, basic LLM agent concepts. Optional: OpenAI API, RAG pipelines.
• Backend: Python, Django, Django REST Framework, PostgreSQL, Celery, Redis, JWT/OAuth2, Docker.
• DevOps/Tools: Azure, Git, CI/CD. Optional: Sentry, AWS.
• Testing & Quality (optional): pytest, TDD.
• Frontend: React/Next.js, TypeScript, REST API integration.

Preferred qualifications

Must-have

• Python + Django/DRF fundamentals: 0–2 years of hands-on work (internships, projects, open source count). Models, serializers, viewsets, migrations.
• SQL & PostgreSQL: relational modelling, joins, indexes at a basic level.
• LLM API experience: built at least one project calling an LLM API (Anthropic preferred) with tool/function calling — shown on GitHub or in a demo.
• Skills, agents & MCP basics: has written at least one agent skill (SKILL.md-style) or custom agent/subagent, and built a simple custom MCP server/connector (Python or TypeScript) exposing a few tools.
• API basics: REST design, JWT/OAuth2 authentication, Docker basics.
• Git discipline: branching, pull requests, resolving conflicts.
• Communication: clear written updates; asks early when blocked; takes review feedback well.

Strongly preferred

• Agentic AI coding tools used day to day, with a view on where generated code needs extra review.
• Custom MCP connector creation: built and tested an MCP server wrapping a real API, with auth and clear tool descriptions.
• Agent skills & custom agents: authored skills with references/scripts, or configured custom agents with scoped tools.
• Celery/Redis on a background-job feature.
• React/Next.js + TypeScript for small UI work.

Nice-to-have: pytest and TDD habits; Azure or AWS basics; Sentry; RAG pipelines; B.E./B.Tech/MCA/BCA in CS or equivalent proof of work.

This is first and foremost a backend + AI-workflow hire.

Weight evaluation toward Python/Django fundamentals, problem solving and real LLM projects — not framework trivia.