Staff AI Engineer, Internal Automation

CargoSprint

1d•Hybrid

About The Position

This is a Staff-level role focused on building the internal automation layer that powers the next chapter of a company moving at speed. The role requires deep technical credibility and organizational influence to identify, prioritize, and build AI-powered automation. The ideal candidate is passionate about solving complex problems with AI, has a strong business sense, and can communicate effectively with both technical and non-technical stakeholders. The role involves owning the full lifecycle of AI agents, from workflow analysis to deployment and iteration, and integrating them with existing systems. Additionally, the engineer will contribute to building the company's internal AI capabilities by documenting work, establishing patterns, and helping the broader engineering team develop agent-building fluency. The role also involves using AI coding tools to accelerate development and potentially traveling to the Guadalajara office to observe workflows and drive adoption.

Requirements

Extreme ownership — you treat what you ship as yours. You do not hand off and walk away.
8+ years of engineering experience, with a meaningful portion spent building systems that internal teams depend on
Track record of shipping AI or automation into production at a company that was moving fast and had high standards
Strong business sense — you understand ROI, you can read a workflow and identify where the real cost is, and you can make the case for your prioritization in business terms
Excellent communication in English — written and verbal; you can explain a complex system to an operations manager and to a staff engineer in the same week without losing either of them
Comfortable working with limited oversight and ambiguous problem statements — you convert ambiguity into a scoped proposal, not a question
Willingness to travel to CargoSprint's Guadalajara, Mexico office as needed to work directly with the teams you are building for
Expert-level Python — you write clean, production-grade code, and your code reviews raise the bar
Deep experience with LLMs and agent frameworks — LangChain, LangGraph, or equivalent; you have built multi-step agentic workflows that run reliably in production, not just in notebooks
API and integration experience — you can connect an agent to any system that exposes an API, and you know what to do when it doesn't
FastAPI and backend service design — you build the infrastructure your agents run on, not just the agents themselves
Postgres and database fluency — schema design, query optimization, vector extensions (pgvector); you know how data should flow through a system
DevOps and deployment — Docker, CI/CD, monitoring and alerting; you ship to production and you know when something is wrong before a user does
RAG and retrieval systems — you have built systems that ground LLM outputs in real organizational knowledge, not just general model capability

Nice To Haves

Familiarity with most of the following technologies: Python, FastAPI, Temporal, PostgreSQL, LangGraph, LLMs (Open AI, Claude), Embeddings, Docker, Kubernetes, Pulumi, Salesforce, Hubspot, REST APIs.

Responsibilities

Identify and prioritize the highest-ROI automation opportunities across CargoSprint's internal departments — sales, finance, customer operations, and beyond
Design and build production-grade AI agents using Python, LLMs, and agent orchestration frameworks that internal teams actually depend on
Own the full lifecycle: from workflow analysis and architecture through implementation, deployment, monitoring, and iteration
Partner with department leaders and the SDR, finance, and operations teams to deeply understand their workflows before touching a keyboard
Instrument everything — build dashboards and telemetry that prove business impact in terms department heads understand (hours saved, process steps eliminated, error rates reduced)
Integrate agents with existing systems: Salesforce, HubSpot, ZoomInfo, Postgres, internal APIs — whatever the department runs on
Build the internal AI muscle: document what you build, establish patterns and reusable components, and help the broader engineering team develop agent-building fluency
Use AI coding tools to accelerate your own development — Cursor, Copilot, Claude — and raise the team's productivity bar alongside yours
Travel to CargoSprint's Guadalajara office as needed to observe workflows firsthand, build relationships with the teams you are automating for, and drive adoption on the ground