Role Overview

We are looking for an **AI-First Android Developer** who builds production-grade Android applications by working alongside AI agents and models — not by writing every line by hand. The right candidate treats tools like Claude, Codex (GPT), and Gemini as primary collaborators for design, implementation, debugging, and shipping. This is a strict **AI-first role**.

Responsibilities

Design and ship production-grade Android apps where the entire codebase is generated, refactored, and maintained through AI agents (Claude, Codex/GPT, Gemini, etc.).
Drive features end-to-end using agentic workflows — break problems into specs, prompt the right model/agent, validate output, and integrate.
Own GitHub-based AI pipelines: CI/CD, automated PR reviews, AI-assisted testing, and agent-driven release workflows.
Use coding agents (Claude Code, Codex CLI, Cursor, Aider, OpenCode, etc.) as the primary IDE — not as autocomplete.
Architect prompts, context, and tool integrations (MCP servers, function calling, structured outputs) so agents can autonomously deliver features.
Debug and fix issues by directing AI agents — never by manual copy-paste patching of bugs.
Integrate REST / JSON / XML APIs and web services through agent-generated, well-tested code.
Design clean, Material-3-compliant UIs and let agents implement them; you review, refine, and ship.
Improve app performance, cold start, memory, and battery using AI-driven profiling and code generation.

Requirements

Must-Have — AI Skills

Strong, hands-on knowledge of AI agents and large language models — what they're good at, where they fail, and how to steer them.
Daily working experience with Claude (Anthropic), GPT (OpenAI), and Gemini (Google) for real production code.
Comfort with agentic coding tools: Claude Code, Codex CLI / OpenCode, Cursor, Aider, Windsurf, or equivalent.
Prompt engineering for code: writing clear specs, providing the right context, using system prompts, structured outputs, and tool/function calling.
Knowledge of MCP (Model Context Protocol) servers and how to wire agents into real tools (GitHub, file system, build systems, test runners).
Familiarity with GitHub-based AI pipelines: GitHub Actions, AI-powered code review (e.g., CodeRabbit, Claude/Copilot reviewers), auto-PR agents, and agent-driven CI/CD.
Ability to evaluate AI output critically — catch hallucinated APIs, insecure code, and bad patterns before they ship.

Must-Have — Android Skills

Solid grounding in Kotlin (preferred) and core Java.
Strong understanding of Android SDK, Jetpack, Coroutines/Flow, and modern architecture (MVVM / MVI, Clean Architecture).
Jetpack Compose and/or XML layouts, with a working knowledge of Material 3 design guidelines.
REST API integration and confident handling of JSON and XML payloads — including parsing, error handling, and retries.
Git and GitHub fluency: branching, PRs, code review, releases, and Actions workflows.
Ability to ship and own a project independently, end-to-end, from spec to Play Store.

How We Work (Read Carefully)

AI-first, always. Code is produced through AI agents. We do not pay engineers to hand-type boilerplate or copy-paste fixes from forums.
No traditional copy-paste debugging. Bugs are reproduced, described, and resolved through agent-driven workflows with proper context and tests.
Every PR is expected to be agent-assisted, well-tested, and reviewed by both a human and an AI reviewer before merge.
You will be measured on shipped quality and velocity — not on lines of code typed.
You are expected to keep learning new models, agents, and tools as the AI ecosystem evolves weekly.

Nice to Have

Experience building or integrating custom MCP servers.
Exposure to on-device ML (TensorFlow Lite, ML Kit, Gemini Nano) or in-app LLM features.
Open-source contributions, especially to AI tooling or Android libraries.
Experience with Kotlin Multiplatform (KMP) or Compose Multiplatform.