Supply chains run the world. We make them run themselves. At Sensos, we turn real-world data into a trusted foundation that enables companies to transform their supply chains into autonomous execution services, acting before problems arise, at scale. We're a fast-moving global team at the intersection of data, agentic AI, autonomous execution, and global supply chains.

As a Data Scientist focused on AI Agents, you will design autonomous agents that monitor real-time supply chain visibility data and execute automated actions directly within Sensos's core platform. You'll own this work end-to-end, from shaping requirements with Product to deploying and evaluating production systems.

Responsibilities:

• Design and deploy agent systems. Architect and ship multi-agent workflows and decision-making logic that monitor end-to-end supply chain logistics and trigger automated, real-time responses.
• Own evaluation and observability. Build the eval frameworks, monitoring, and safe-rollout practices that let us trust autonomous agents in production - measuring accuracy, drift, and real-world impact against supply chain outcomes.
• Translate product needs into systems. Partner directly with Product Management to turn business logic, user needs, and logistics metrics into scalable analytical and agentic solutions.
• Build with engineering. Work alongside backend engineers to integrate real-time IoT telemetry and streaming pipelines into agent-driven actions.
  
  

Requirements:

• BSc or MSc in Computer Science, Data Science, or related field (research experience is an advantage).
• 3+ years in Data Science, Machine Learning, or a closely related field.
• GenAI in production. Proven track record building production GenAI applications: multi-step agents, RAG pipelines, tool-augmented LLMs, and structured outputs.
• Strong evaluation and observability practices for LLM systems - this is a must. You've designed eval sets, run regression tests on prompts and agents, monitored production LLM behavior, and made safe-rollout decisions based on real metrics.
• Agent orchestration. Hands-on experience designing and orchestrating multi-step or multi-agent systems - managing state, tool use, retries, fallbacks, and handoffs between agents.
• Voice agents. Experience building conversational voice agents that handle real-time phone or voice interactions.
• Strong ML fundamentals. Solid grounding in classical ML and deep learning - model selection, training, evaluation metrics, handling imbalanced data, feature engineering, and the trade-offs between model families. Comfortable reasoning about when an LLM is the right tool and when it isn't.
• Hands-on with the modern agent stack. Proficient in Python. Familiarity with major LLM providers like OpenAI, Azure OpenAI, Google Vertex AI, Anthropic, etc.
• Startup mindset. Comfortable working independently in a fast-paced environment, owning ambiguous problems, and finding creative paths forward.
  
  

Nice to have

• Hands-on experience deploying models on Microsoft Azure (e.g., Azure Foundry, Azure AI Search).
• Background in supply chain, logistics, IoT, or other real-time/streaming domains.