AI Developer - PGR

About the role

Aspire Software is seeking an AI Developer to join its Lebanon team. You will be responsible for delivering AI‑driven product features while shaping the architectural roadmap for future capabilities. The role blends hands‑on development with strategic product thinking.

Key responsibilities

• Design, build, and ship AI features using the PHP/MySQL/React/Node stack with OpenAI (or similar) API integration.
• Extend Astro Core chatbot with new read/write actions, graph generation, and analytical outputs.
• Improve the external MCP connector schema, authentication mechanisms, and capability surface.
• Own prompt engineering for product features, documenting prompts as first‑class artifacts.
• Implement token metering and usage‑tracking per feature, customer, and tier.
• Design and enforce guardrails, access controls, and rate‑limiting across the Astro AI platform.
• Contribute to agent architecture decisions, including tool access, memory design, token pool isolation, and training data boundaries.
• Monitor AI feature performance in production (latency, cost, quality) and drive continuous improvements.
• Stay current on AI tooling trends and provide concrete recommendations.

Required profile

• Proven experience integrating LLM APIs (OpenAI, Anthropic, or equivalents) into production applications used by real customers.
• Strong fluency in prompt engineering, including system prompts, few‑shot examples, structured outputs, and tool/function calling.
• Understanding of token economics, cost management, and model selection trade‑offs.
• Experience building quality evaluation and output validation frameworks for AI features.
• Awareness of AI‑specific risks such as hallucination, prompt injection, and data leakage.
• Background in designing metered usage systems, API rate limiting, and billing integration.
• Familiarity with multi‑tenant access controls and agent architecture patterns.

Required skills

• PHP
• MySQL
• React
• Node.js
• OpenAI API (or similar LLM APIs)
• Prompt engineering
• Token metering and usage tracking
• Rate limiting and access control implementation
• AI performance monitoring
• Agent architecture concepts (tool registries, memory types, token pool management)