Deion is a done-for-you, anti-AI-slop Canva alternative for YouTube thumbnail design, powered by a proprietary algorithm. They are seeking an undergraduate Agentic AI Engineering Intern to design and implement custom autonomous agent architectures and collaborate with cross-functional teams in a cloud-native environment.
Responsibilities:
- Design and implement custom autonomous agent architectures using Python and Google Cloud infrastructure
- Leverage Gemini CLI and Claude Code to accelerate development cycles and orchestrate complex agentic task execution
- Build bespoke reasoning loops and decision-making logic to solve open-ended problems without relying on pre-built low-code platforms
- Integrate diverse data sources and external services into a cohesive system via robust, Python-based middleware
- Debug and refine non-deterministic agent behaviors to ensure reliability, safety, and alignment with user goals
- Architect modular systems where agents can autonomously select and use the correct tools for a given task
- Collaborate with cross-functional teams to define, implement, deploy, and scale AI solutions within a cloud-native environment
- Document system designs, source code, workflow designs, research, development plans, and engineering standards to enable solution modularity, reproducibility, portability, collaboration, and knowledge sharing
- Collaborate with teammates using Slack, Notion, Google Meet, and/or other software
Requirements:
- Completion of a demo project
- At least 10 hours of synchronous meeting availability between 1:00 PM and 5:00 PM Central Time on weekdays for the next 6 months
- Comfortable working 100% remotely as part of a distributed team
- Excel at designing autonomous systems that can reason through complex tasks using LLMs
- Work products demonstrate a high level of attention to detail and the ability to see the big picture
- Aptitude, mindset, and ability to learn quickly
- Design and implement custom autonomous agent architectures using Python and Google Cloud infrastructure
- Leverage Gemini CLI and Claude Code to accelerate development cycles and orchestrate complex agentic task execution
- Build bespoke reasoning loops and decision-making logic to solve open-ended problems without relying on pre-built low-code platforms
- Integrate diverse data sources and external services into a cohesive system via robust, Python-based middleware
- Debug and refine non-deterministic agent behaviors to ensure reliability, safety, and alignment with user goals
- Architect modular systems where agents can autonomously select and use the correct tools for a given task
- Collaborate with cross-functional teams to define, implement, deploy, and scale AI solutions within a cloud-native environment
- Document system designs, source code, workflow designs, research, development plans, and engineering standards
- Collaborate with teammates using Slack, Notion, Google Meet, and/or other software
- Strong proficiency in Python with a focus on writing clean, scalable, and highly modular code
- Practical experience deploying and managing applications within the Google Cloud ecosystem
- Familiarity using advanced developer tools like Gemini CLI or Claude Code to solve technical challenges
- Demonstrated systems thinking skills with the ability to map out intricate logic flows and manage state across complex processes
- Ability to decompose high-level, ambiguous objectives into a series of small, autonomous, and verifiable tasks
- Experience using Git-based version control to manage codebases within a collaborative team setting
- Excellent communication skills for explaining technical trade-offs and architectural decisions to both technical and non-technical peers
- A high degree of adaptability and the 'engineer's intuition' required to troubleshoot probabilistic systems
- Effective modular thinking skills evidenced by the ability to break down large, intricate undertakings into a series of smaller self-contained tasks
- Excellent written and verbal communication skills
- Emotional intelligence and coachability