Salary
💰 $130,000 - $160,000 per year
Tech Stack
CloudPythonReact
About the role
- Design, build, and optimize agentic AI systems that power Selector’s operational intelligence platform.
- Leverage emerging frameworks (e.g., pydantic-ai, LangChain) and evaluate new agentic AI technologies to accelerate development and maintain cutting-edge capabilities.
- Develop empirical pipelines for measuring and assessing agent performance, ensuring all prompt and model changes are backed by data-driven evidence and quantifiable improvement.
- Implement advanced Natural Language Processing (NLP) techniques to translate natural language queries into complex structured queries against operational data sources.
- Design human-in-the-loop workflows for error detection, correction, and refinement—enabling agents to prompt for clarifications when necessary.
- Build transparent tracking mechanisms for agent thoughts, decisions, and observations throughout iterative task flows to improve interpretability, debugging, and trust.
- Evaluate and apply different agent paradigms (ReAct, reflex, goal-based, utility-based, etc.) to align agent behavior with specific task requirements.
- Collaborate cross-functionally with data engineers, product managers, and customer success teams to ensure AI-driven features align with customer needs and business outcomes.
- Contribute to the evolution of Selector’s conversational UX, making agent interactions more natural, reliable, and contextually aware.
Requirements
- Bachelor’s or Master’s degree in Computer Science, Data Science, or related technical field; or equivalent practical experience.
- 3-5 years of software engineering experience, with at least 1 year of experience building Agentic AI systems.
- Strong programming skills in Python and familiarity with modern AI/ML ecosystems.
- Experience building or integrating agent-based systems using frameworks such as pydantic-ai, LangChain, or similar.
- Solid understanding of Natural Language Processing and prompt engineering techniques.
- Proven ability to design metrics-driven evaluation pipelines for AI/LLM performance testing.
- Knowledge of agent reasoning strategies (e.g., ReAct, reflexive, goal-driven, utility-based) and practical experience choosing among them.
- Familiarity with human-in-the-loop systems, error handling, and recovery strategies.
- Strong problem-solving, analytical, and debugging skills, with attention to reproducibility and system robustness.
- Excellent communication skills, with the ability to collaborate in a fast-paced startup environment.
- Bonus: Experience with operational data domains (networking, cloud, application performance) or conversational UX design.