Agent Engineer (Builder)

Omninet Capital•Beverly Hills, CA

10d•$100,000 - $145,000•Onsite

About The Position

We're looking for a builder, not just a coder — someone who understands agentic AI well and is excited to apply it to real operational problems in a real estate investment and property management platform. You'll partner directly with department heads and property teams to understand repetitive, manual, or error-prone workflows, take the business rules and parameters from them, then design and ship AI agents and automations that eliminate them. You don't need to be an expert in leasing or accounting, and you don't need to build full code backends: most of the work is prompts, skills, and low-code automation. This is a highly cross-functional, high-visibility role reporting into IT, with exposure to Acquisitions, Accounting, Property Management, HR, and Leasing.

Requirements

2+ years of experience in automation, business systems, solutions engineering, or a similar technical role; a degree in Computer Science or a related field is a plus but not required
Strong hands-on experience with LLMs and agentic tools, including writing prompts, skills, or instructions that produce consistent, reliable results
Experience building skills, custom assistants, plugins, or connectors on AI platforms such as Claude, ChatGPT, Microsoft Copilot, or Grok
Familiarity with APIs and basic automation concepts
Strong written and verbal communication skills — able to translate business problems into technical solutions and explain technical solutions to non-technical stakeholders
Self-directed, comfortable with ambiguity, and eager to work directly with business teams (not just other engineers)
Good judgment about data sensitivity and when a person must review AI output
Willingness to travel occasionally

Nice To Haves

Experience with low-code/no-code workflow tools and the Microsoft 365 ecosystem
Comfortable reading and adjusting code (Python, PowerShell, JSON) when needed, including code written with AI assistance
Familiarity with Model Context Protocol (MCP) or API-based agent development
Exposure to real estate, property management, or financial services
Familiarity with Yardi or similar property management systems

Responsibilities

Interview stakeholders across Acquisitions, Accounting, Operations, HR, and Leasing to map current manual workflows
Design, build, and deploy AI agents (skills, plugins, and connectors on our AI platforms, plus rules-based automations) that automate document review, data entry, reporting, underwriting support, and internal Q&A
Integrate agents with existing systems (Microsoft 365, SharePoint, Outlook, Yardi, Rippling) using available connectors and low-code/no-code tools
Package finished tools so non-technical users can run them easily, with short how-to guides
Work with regional and on-site property management teams to identify high-friction tasks (maintenance ticketing, resident communications, lease abstraction, vendor invoicing, compliance tracking)
Deploy automation and AI agents that reduce manual data entry and turnaround time across the ~8,500-unit multifamily portfolio and commercial assets
Travel occasionally to properties when on-site requirements gathering or rollout support is needed; most discovery happens remotely over Teams
Monitor, maintain, and iterate on deployed agents; track efficiency gains (hours saved, error reduction, cost avoidance)
Document workflows and maintain a library of reusable automation components
Test each automation with its business owner before release, and keep a person in the loop for anything sent outside the company
Work within IT's security model: least-privilege access, approved connectors, and service accounts
Stay current on AI/agent tooling and recommend new use cases, including when a workflow is ready for more advanced, API-based agents
Deliver several working automations for real teams, such as lease renewal letter drafting and broker deal intake
Measure results (hours saved, turnaround time, error rates) and share what worked and what didn't
Build a reusable library of skills, templates, and connectors that makes each new automation faster
Use what we learn together to shape the program's next stage