• Develop novel reinforcement learning algorithms to solve complex, real-world problems. • Integrate Large Language Models (LLMs) to enhance agent reasoning, threat intelligence, or human-machine interactions. • Implement and extend Message Control Protocol (MCP) to enable coordinated behavior across modular AI components or multi-agent systems. • Prototype new methods and help transition successful prototypes into deployed solutions. • Collaborate with cross-functional teams (Data Science, Engineering, Product) to integrate and deploy RL models in production. • Conduct thorough evaluations of model performance using appropriate experimental design and statistical analysis. • Write clean, maintainable and well-documented code. • Participate in code reviews and ensure the code quality of your team. • Troubleshoot and debug complex software issues.

• Own the complete AI platform portfolio across three categories: AI-enabled UX, AI agents (MCP/agentic automation), and AI in data processing. • Drive embedded AI features from internal alpha through production — managing rollout criteria, prompt iteration, success rate thresholds, and the decision framework for when each feature is ready for customer availability. • Define and maintain Redox's AI investment framework: KPIs, rollout criteria, cost guidelines, and the research agenda for what we still need to learn (monitoring, agentic loops, RAG, re-prompting). • Manage beta relationships for agentic use cases across multiple customer segments. • Lead Redox's agentic product strategy — starting with the MCP Server as the current foundation, expanding into native agent support for high-value customer workflows throughout 2026, with a clear 2027 vision & roadmap. • Equip sales and CS to position Redox's AI platform story accurately. • Scope and drive overall platform features required to unlock new use case, channel partnerships, & commercial success in the Enterprise customer base. • Ensure complex infrastructure work (VPN configurations, multi-tenant patterns, integration partner requirements) is scoped clearly and investments are made with commercial need. • Partner with GTM & Customer Success to enable vendor customer growth (both new business and expansion). • Contribute to use case development, customer segmentation, & field enablement to close high-value Enterprise vendor deals.

• Lead the design and implementation of production AI systems that solve real business problems for our partners • Spot market opportunities to match technology with business outcomes, bringing a builder's instinct for what will and won't land commercially • Drive Able's AI strategy forward by constantly exploring and implementing cutting-edge AI technologies, frameworks, and tools • Work with partners and prospects to understand their business and conceive of opportunities to introduce technology (AI specifically) to improve performance • Provide coaching and mentorship, continuously enhancing our team's and partner teams AI capabilities through knowledge sharing and training on the latest technology and best practices

• Bridge the gap between theoretical AI research and practical business applications by building end-to-end LLM-powered features. • Optimize retrieval-augmented generation (RAG) pipelines and ensure the reliability of AI services in a cloud-native environment. • Design and deliver production applications using model APIs (OpenAI, Anthropic, Gemini) and orchestration frameworks like LangChain, LlamaIndex, or LangGraph. • Build and optimize retrieval systems over proprietary data using vector databases such as Pinecone, Weaviate, or Milvus and hybrid search techniques. • Develop autonomous agents and multi-step reasoning workflows that call external tools and maintain state to solve complex automation tasks. • Establish evaluation pipelines (using frameworks like DeepEval) to measure model drift, accuracy, and latency, ensuring safe and ethical AI outputs. • Package AI applications in Docker containers and manage scalable deployments on cloud platforms (AWS, Azure, or GCP) using CI/CD pipelines. • Design pipelines for data ingestion, cleaning, and chunking to support retrieval and model fine-tuning.