Apply on Employer Site

Downstream · 13 hours ago

Senior Applied AI Engineer (Agentic Systems & MCP)

United States

Full-time

Remote

Senior Level

$140K/yr - $165K/yr

Downstream Systems, Inc. is a leading marketplace platform transforming the waste and construction industries. They are seeking a Senior Applied AI Engineer to architect the intelligence layer of their marketplace, integrating AI into their platform and building robust systems for complex logistics problems.

ConstructionLogisticsMarketplaceRecyclingWaste Management

Hiring Manager

Lauren Geber

Responsibilities

Architect MCP Servers: Design and implement Model Context Protocol (MCP) servers to standardize how our data (inventory, logistics, pricing) is exposed to AI agents. You will treat context as a first-class product

Agentic Orchestration: Build the backend logic for multi-agent systems that can plan, reason, and execute tasks. You will ensure these agents can plug into client-facing Chat UIs (built in Flutter/React) via structured, reliable protocols

LLM-Powered Recommendation Engines: Develop predictive engines that analyze historical construction data and project requirements to proactively recommend rentals and services

Rigorous LLM Evaluation (LLMOps): You will implement comprehensive logging and evaluation pipelines using tools like Weights & Biases (WandB) . You are responsible for measuring drift, hallucination rates, and agent success metrics— ensuring we ship reliable intelligence, not just "vibes."

Production Python Backend: Write high-performance, maintainable Python code. Whether it's a microservice or a monolith, you ensure the AI logic is scalable and secure

Qualification

Senior Applied AI ExperienceMastery of the AI StackMCP ExpertisePython FundamentalsEvaluation-First MindsetDjango ExperienceLogistics Domain KnowledgeMobile/Frontend Awareness

Required

Senior Applied AI Experience: You have moved beyond prototypes. You have experience deploying LLM applications where reliability and latency matter

Mastery of the AI Stack: Deep expertise with orchestration frameworks (LangChain, LangGraph, or custom implementations) and vector databases

MCP Expertise: You understand the Model Context Protocol and how to build servers that provide context to LLMs efficiently

Evaluation-First Mindset: You don't guess if a prompt is better; you prove it. Extensive experience with logging and evaluation platforms (WandB, LangSmith, Ariz) is required

Strong Python Fundamentals: You are an expert in Python. You understand concurrency, async/await patterns, and API design (REST/gRPC)

Builder's DNA: You are comfortable with ambiguity and iteration. You build scrappy to validate, but architect for scale

Preferred

Django Experience: Our core platform is built on Django. Familiarity is helpful, but strong Python fundamentals are what matter

Mobile/Frontend Awareness: Experience working with Flutter or modern JS frameworks. You don't need to write the UI, but you need to understand how to feed it

Logistics Domain Knowledge: Familiarity with marketplaces, supply chain, or construction tech