C
AI-Factory/Harness Engineer (Remote - LATAM, Canada, US)
Accepting applicationsCloudGeometry · United States
Full-Time Mid_senior AIAiPythonaiasic
Posted
27 May
Category
Test
Experience
Mid_senior
Country
United States
AI Lifecycle Engineer
(AI Harness & Factory Engineer / Supervisor)
CloudGeometry is building AI-MSL — an end-to-end, AI-powered, and supervised software development and maintenance platform that transforms how production systems are maintained, modernized, and extended. AI-MSL combines semantic system intelligence (AppGraph), AI-driven SDLC execution, and expert-supervised lifecycle governance to replace coordination-heavy, traditional development models.
We are seeking an AI-Native Engineer to serve as a Harness and Factory Developer and Supervisor. In this role, you will be responsible for building, optimizing, and monitoring the automated AI factory workflows and low-level agent harnesses that drive autonomous code generation. You are deeply fluent with modern AI engineering, have a strong engineering foundation in TypeScript and Node ecosystems, and actively contribute to the broader open-source community.
Key Responsibilities
AI Harness & Factory Engineering
Harness & Skills Development: Design, build, and optimize low-level AI coding harnesses, writing custom skills, execution hooks, and integration layers to control model behavior.
Factory Supervision: Supervise automated code-generation factory pipelines, ensuring quality, trace integrity, and deterministic reliability of output.
Workflow Optimization: Build and refine low-level AI orchestration patterns using graph-based agent frameworks to model complex engineering workflows.
Open Source Contribution: Actively participate in, submit to, and collaborate with upstream Open Source Software (OSS) projects related to AI coding harnesses and agent infrastructure.
Core Backend & Data Operations
Systems Engineering: Build and extend robust, asynchronous platform components using Nest.JS, TypeScript, and React.
Data & Queue Management: Manage high-scale transactional data and background processing workloads using PostgreSQL, pg-boss, and Redis.
Event-Driven Architecture: Design and implement decoupled workflows following Event-Driven Architectures (EDA) and microservices patterns.
Infrastructure Parity: Containerize services using Docker and docker-compose to guarantee smooth execution across local, CI, and cloud environments.
Technical Requirements & Qualifications
AI Engineering & Orchestration
Low-Level Orchestration: Hands-on experience building multi-agent or stateful workflows using LangGraph or equivalent graph-based LLM frameworks.
AI Tool Fluency: Practical mastery of ClaudeCode, Codex, or Opencode engineering patterns.
Harness Modification: Proven ability to write low-level code hooks, telemetry interception, and prompt context-grounding strategies.
Core Stack & Software Engineering
Languages & Frameworks: Strong professional background in TypeScript, Nest.JS, and React. Broad language versatility (e.g., Python or polyglot experience) is highly valued.
Databases & Event Tooling: Production experience with PostgreSQL, database-backed job queues (pg-boss), and Redis.
Architecture & Cloud: Good conceptual or basic practical grasp of microservices and Event-Driven Architectures (EDA).
Cloud Infrastructure: Grounding in core AWS Services, specifically: S3, CloudWatch, RDS, EC2, SES, SNS, and Lambdas.
OSS & Deployment Experience
OSS Participation: Track record of active participation, pull request submissions, or issue governance in open-source software repositories.
Startup Mindset: A strong ownership mentality with customer-facing comfort, ready to translate loose requirements into automated code factory specifications.
Preferred AI Factories Experience
We favor candidates who have actively broken down, extended, or built on top of public and private code-generation loops. Experience with the following is highly preferred:
OpenHands, SWE-agent, Aider, Devin-style autonomous coding workflows, OpenDevin, or Continue.dev.
Custom ClaudeCode orchestration layer development and semantic code graph systems.
Why CloudGeometry?
CloudGeometry is moving away from fragmented, plug-in based chat models to build enterprise-ready, governed software manufacturing factories. As a core developer and supervisor of these loops, your engineering contributions will directly change how software is authored, shipped, and maintained globally.
Comprehensive compensation and benefits package.
Extensive training, hackathons, and digital learning platforms (Udemy).
Developer Pro access to ClaudeCode, Codex, and AntiGravity.
Fully collaborative environment working alongside global experts with zero legacy systems.
Show more Show less
(AI Harness & Factory Engineer / Supervisor)
CloudGeometry is building AI-MSL — an end-to-end, AI-powered, and supervised software development and maintenance platform that transforms how production systems are maintained, modernized, and extended. AI-MSL combines semantic system intelligence (AppGraph), AI-driven SDLC execution, and expert-supervised lifecycle governance to replace coordination-heavy, traditional development models.
We are seeking an AI-Native Engineer to serve as a Harness and Factory Developer and Supervisor. In this role, you will be responsible for building, optimizing, and monitoring the automated AI factory workflows and low-level agent harnesses that drive autonomous code generation. You are deeply fluent with modern AI engineering, have a strong engineering foundation in TypeScript and Node ecosystems, and actively contribute to the broader open-source community.
Key Responsibilities
AI Harness & Factory Engineering
Harness & Skills Development: Design, build, and optimize low-level AI coding harnesses, writing custom skills, execution hooks, and integration layers to control model behavior.
Factory Supervision: Supervise automated code-generation factory pipelines, ensuring quality, trace integrity, and deterministic reliability of output.
Workflow Optimization: Build and refine low-level AI orchestration patterns using graph-based agent frameworks to model complex engineering workflows.
Open Source Contribution: Actively participate in, submit to, and collaborate with upstream Open Source Software (OSS) projects related to AI coding harnesses and agent infrastructure.
Core Backend & Data Operations
Systems Engineering: Build and extend robust, asynchronous platform components using Nest.JS, TypeScript, and React.
Data & Queue Management: Manage high-scale transactional data and background processing workloads using PostgreSQL, pg-boss, and Redis.
Event-Driven Architecture: Design and implement decoupled workflows following Event-Driven Architectures (EDA) and microservices patterns.
Infrastructure Parity: Containerize services using Docker and docker-compose to guarantee smooth execution across local, CI, and cloud environments.
Technical Requirements & Qualifications
AI Engineering & Orchestration
Low-Level Orchestration: Hands-on experience building multi-agent or stateful workflows using LangGraph or equivalent graph-based LLM frameworks.
AI Tool Fluency: Practical mastery of ClaudeCode, Codex, or Opencode engineering patterns.
Harness Modification: Proven ability to write low-level code hooks, telemetry interception, and prompt context-grounding strategies.
Core Stack & Software Engineering
Languages & Frameworks: Strong professional background in TypeScript, Nest.JS, and React. Broad language versatility (e.g., Python or polyglot experience) is highly valued.
Databases & Event Tooling: Production experience with PostgreSQL, database-backed job queues (pg-boss), and Redis.
Architecture & Cloud: Good conceptual or basic practical grasp of microservices and Event-Driven Architectures (EDA).
Cloud Infrastructure: Grounding in core AWS Services, specifically: S3, CloudWatch, RDS, EC2, SES, SNS, and Lambdas.
OSS & Deployment Experience
OSS Participation: Track record of active participation, pull request submissions, or issue governance in open-source software repositories.
Startup Mindset: A strong ownership mentality with customer-facing comfort, ready to translate loose requirements into automated code factory specifications.
Preferred AI Factories Experience
We favor candidates who have actively broken down, extended, or built on top of public and private code-generation loops. Experience with the following is highly preferred:
OpenHands, SWE-agent, Aider, Devin-style autonomous coding workflows, OpenDevin, or Continue.dev.
Custom ClaudeCode orchestration layer development and semantic code graph systems.
Why CloudGeometry?
CloudGeometry is moving away from fragmented, plug-in based chat models to build enterprise-ready, governed software manufacturing factories. As a core developer and supervisor of these loops, your engineering contributions will directly change how software is authored, shipped, and maintained globally.
Comprehensive compensation and benefits package.
Extensive training, hackathons, and digital learning platforms (Udemy).
Developer Pro access to ClaudeCode, Codex, and AntiGravity.
Fully collaborative environment working alongside global experts with zero legacy systems.
Show more Show less