Senior+ Software Engineer (AI Agents)

tryopine.com•Durham, NC

5h•Onsite

About The Position

Opine is a fast-growing AI sales tech startup trusted by leading companies like Docker, Socket Security, BigID, and MaintainX. Opine transforms scattered conversations, notes, and documents into dynamic, living account plans. The platform proactively surfaces risks, aligns every stakeholder, and automates repetitive work from discovery to delivery. The result is faster sales cycles, cleaner handoffs, and consistent execution across sales, presales, and customer success. The founding team were all early employees at cybersecurity unicorn, JupiterOne, and led key engineering and sales teams that led to its rapid growth and 8 figure ARR in 2 years. As a software engineer at Opine focused on AI agents, you’ll help shape how our AI systems think, plan, and take action in the real world. You'll be building the core agent architecture, including multi-step planning, memory, and tool integration, so our agents can reliably handle complex workflows. You'll be working at the cutting edge, building agents that interact with massive volumes of sales data from some of the most forward-thinking companies.

Requirements

Significant experience building and maintaining production AI agents powered by modern language models, handling complex, real-world use cases.
Significant experience working with relational databases like PostgreSQL, and vector databases like pgvector, Pinecone, or Weaviate.
Experience working with data infrastructure, including technologies like Delta Lake or analytics systems like ClickHouse.
Strong background building with Node.js or Python.
Ability to navigate across the full cycle of development.
Ability to own a difficult problem in its entirety.
Care about speed and working with extremely high performers.

Nice To Haves

Experience managing cloud infrastructure on AWS, Azure, or GCP.
Experience using tools like Terraform.
Affinity for writing.

Responsibilities

Building the core agent architecture, including multi-step planning, memory, and tool integration.
Ensuring agents can reliably handle complex workflows.
Building agents that interact with massive volumes of sales data.
Navigating across the full cycle of development, from idea/customer conversation to design of the backend system, design of the frontend experience, and management of the infrastructure.
Owning a difficult problem in its entirety.
Building and shipping evaluation pipelines and datasets to benchmark and improve agent performance in production.