Senior Software Engineer, AI Platform

About Mirai and Role Overview

Mirai, a Scopely Studio based in Riyadh, is dedicated to crafting compelling gaming experiences enjoyed by millions globally. We are seeking a Senior Software Engineer specializing in AI platform development, who is adept at both product feature creation and designing scalable shared services that underpin multiple products. The AI Native Engineering team develops software tools that leverage large language models (LLMs) for various applications like bug triaging, player feedback analysis, market research, and contract analytics, alongside direct-to-consumer storefronts for Scopely games. Our current challenge includes standardizing model access, permission control, and safety patterns across products.

Responsibilities

• Develop end-to-end product features for internal AI applications using TypeScript or Python, from initial conception through deployment.
• Identify recurring functionalities across individual products and consolidate them into shared services and libraries to facilitate reuse without disrupting ongoing delivery.
• Create MCP tool servers that enable controlled exposure of corporate data respecting user permissions before any model interaction.
• Implement relationship-based access control systems to unify and streamline access rules applied consistently to database queries and vector searches.
• Standardize LLM model access and cost management via a centralized gateway to prevent duplication across repositories.
• Disseminate safety mechanisms such as prompt injection mitigation, kill switches, sandboxed code execution, and tightly controlled data access from isolated projects to all products.
• Ensure shared components are easy to adopt through clear APIs, versioning, example templates, and thorough documentation.
• Lead by example in code reviews, including AI-assisted coding, mentoring engineers during review processes, pair programming, and design discussions.
• Take ownership of incidents related to your work and collaborate with DevOps teams on deployment and operational reliability.

Requirements

• A minimum of six years’ experience in production software development, including a track record of delivering shared services, internal APIs, or libraries relied upon by multiple teams.
• Proficiency in TypeScript and Node.js, and strong skills in Python development.
• Hands-on experience deploying LLM features to production environments including handling embeddings, tool integration, structured outputs, and output quality validation.
• Expertise in designing APIs or libraries with an emphasis on version management, backward compatibility, secure defaults, and comprehensive documentation.
• Experience in building multi-tenant or permission-sensitive software that rigorously enforces data access restrictions to prevent unauthorized data or model exposure, treating all model outputs and retrieved data as potentially unsafe inputs.
• Robust system design capabilities covering service boundaries, data modeling, fault tolerance, idempotency, and cost efficiency.
• Practical deployment experience on AWS using Infrastructure as Code tools like Terraform to oversee service delivery in conjunction with DevOps teams.
• Commitment to quality by integrating testing into development workflows and treating continuous integration checks as mandatory gates.
• Excellent communication skills in English for effective technical explanation and stakeholder engagement.
• Familiarity and judicious use of AI coding assistance, with the ability to fully understand and own merged code outputs.

Technology Stack

• Languages: TypeScript, Node.js, Python 3.11
• Frameworks: Express, FastAPI, Next.js
• AI Tools: OpenAI models and embeddings, LangChain, and relationship-based access control via SpiceDB (in design and rollout)
• Internal Tools: Model Context Protocol (MCP) SDK and LLM gateway built on LiteLLM over Amazon Bedrock
• Cloud & Delivery: AWS services including EC2, ECS Fargate, Lambda, App Runner, S3, CloudFront, Terraform, and GitLab/GitHub
• Data Technologies: PostgreSQL (Amazon RDS and Aurora), BigQuery, Amazon OpenSearch Serverless for vector search
• Identity & Access: Okta, OAuth 2.0 with PKCE, relationship-based access control with SpiceDB
• Project Management: Jira, Confluence

Preferred Qualifications

• Hands-on experience with relationship-based access control solutions like SpiceDB or OpenFGA.
• Familiarity with LLM gateways such as LiteLLM or Amazon Bedrock, including evaluation or tracing tools like Braintrust, Langfuse, or LangSmith.
• Experience with agent orchestration frameworks such as LangGraph.
• Background in maintaining shared platforms serving multiple teams or departments.
• Exposure to gaming industry or Arabic language applications.

Employee Benefits

• Mobile phone allowance
• Comprehensive health insurance coverage
• Performance-based bonus scheme
• Relocation support when applicable