- Design, build, and deploy agentic workflows (multi-step LLM chains with tool calling, retrieval, and structured output) for real-time, business-critical use cases.
- Engineer for determinism and consistency by implementing constrained decoding, structured outputs, caching layers, and evaluation harnesses.
- Build and maintain evaluation and regression frameworks — automated pipelines that measure accuracy, latency, and behavioral consistency across prompt and model changes.
- Integrate LLM agents with external tools and APIs (databases, rules engines, business systems) using frameworks like LangFuse, LangChain, LangGraph, CrewAI, or custom orchestration.
- Deploy agentic systems on cloud infrastructure (AWS, Azure, and/or GCP), optimizing for low-latency inference and cost efficiency.
- Implement guardrails, fallback logic, and observability to ensure agents fail gracefully and every decision is traceable.
- Collaborate with data scientists, software engineers, and business stakeholders to translate business rules into agent behavior and tool definitions.
- Stay current with the latest advancements in AI agents, large language models, and cloud technologies.
- Practical, hands-on experience building and deploying agentic AI systems in production environments.
- Proficiency in Python and experience building production backend systems.
- Experience with LLM APIs (OpenAI, Anthropic, etc.) and agentic frameworks (LangFuse, LangChain, LangGraph, CrewAI, AutoGen, or equivalent).
- Strong understanding of prompt engineering for reliability: structured outputs, few-shot patterns, chain-of-thought, and techniques that minimize hallucination.
- Experience building evaluation and testing pipelines for AI systems, including behavioral evals and golden-set testing.
- Expertise in at least one major cloud provider (AWS, Azure, and/or GCP).
- Familiarity with Databricks, including experience working with its data engineering and analytics capabilities.
- Familiarity with vector databases (Pinecone, Weaviate, pgvector) and retrieval-augmented generation (RAG) patterns.
- Solid knowledge of version control systems (e.g., Git) and CI/CD pipelines.
- Strong problem-solving skills and ability to work collaboratively across teams.
Preferred Expertise:
- Advanced degree (Master's or PhD) in Computer Science, Machine Learning, or a related field.
- Expertise in containerized deployment with Docker.
- Experience building systems where AI outputs feed directly into business-critical decisions.
- Experience in the transportation and logistics industry.
- Familiarity with MLOps/LLMOps tooling.
- Experience with fine-tuning or distillation to optimize for speed and cost at inference time.
- Knowledge of rules engines or constraint solvers and how to combine them with LLM reasoning.
Top Skills
What We Do
We work with global corporate partners to identify the most pressing challenges that they, and broader society, face. Inspired by these complex problems, we launch startups built by proven entrepreneurs, product leaders and technologists that use their agility and talent to develop transformative solutions. After these companies have matured and proven market fit, our corporate partners are able to acquire them, reaping strategic value while enriching their culture and core business. We believe this to be the shortest road to a faster, cleaner, safer, and more accessible future.
Why Work With Us
We launch and innovate 6-8 portfolio organizations a year where no day is the same.
Gallery







