AI Architect

I’ve interpreted “SSM agents” as social media management agents.

AI Architect / Senior AI Automation Engineer
Build and Manage Multiple AI Agents on Local Hardware — Remote

Work arrangement: Fully remote is welcome.
Scope: Architecture, hands-on development, deployment, integration, and ongoing improvement of multiple AI agents.
Compensation: Based on experience, demonstrated technical ability, and agreed scope.

About the Role

We are seeking a hands-on AI Architect / Senior AI Automation Engineer to design, build, deploy, and maintain multiple AI agents on our existing local AI machines.

Our goal is to develop a growing collection of agents that reduce repetitive work, improve efficiency, and optimize workflows across customer communications, marketing, inventory, administration, and other business operations.

This is not a single-chatbot project. You will establish the underlying infrastructure and create multiple specialized agents that can work independently, integrate with our business systems, and coordinate when appropriate.

We are looking for someone who can personally implement and troubleshoot the systems—not only provide recommendations, write prompts, or configure demonstrations.

Agents You Will Build

The initial scope includes the following, with additional agents developed as we identify and prioritize workflow opportunities:

| Agent type | Expected responsibilities |

| AI receptionist | Answer incoming calls, respond to approved questions, capture information, assist with scheduling, route calls, create follow-up tasks, and transfer conversations to a human when necessary. |
| SEO agents | Support keyword research, website audits, content planning, on-page optimization, performance reporting, and identification of opportunities to improve search visibility. Website changes and publication must follow defined approval rules. |
| Social media management agents | Plan content calendars, draft posts and responses, coordinate approved publishing, monitor engagement, and produce performance summaries. |
| Inventory management agents | Monitor inventory records, track usage, identify discrepancies, flag low stock, forecast replenishment needs, and prepare reorder recommendations or approved purchasing workflows. |
| Data entry agents | Extract, validate, organize, and enter information from documents, email s, forms, and spreadsheets into approved systems, with duplicate detection and human review for uncertain information. |
| Additional workflow agents | Automate reporting, document processing, internal knowledge searches, task coordination, reminders, customer follow-up, and other repetitive operational processes. |

The work will be delivered in prioritized phases. You should be able to identify where an AI agent adds value and where a simpler, rules-based automation is more appropriate.

Mandatory Requirement: Local Hosting

All agents must be hosted on our company-owned local AI machines—not on a contractor’s computer or a cloud-hosted agent platform.

The architecture must support local model inference, agent execution, workflow orchestration, and the agents’ supporting data stores. For voice applications, the proposed design should include locally hosted speech processing.

Approved integrations may still connect to telephone carriers, websites, social media platforms, email , calendars, and other business systems. These external services are integration endpoints, not substitutes for hosting the agents and their AI processing locally.

All external dependencies and data transfers must be disclosed and approved. Undisclosed cloud-model calls or automatic cloud-model fallback are not acceptable.

Key Responsibilities

Assess the hardware and design the architecture. Review our existing machines, identify capabilities and limitations, recommend suitable models and software, and plan how multiple agents will share resources. Measure performance and identify any hardware upgrades needed before expanding the system.

Build and integrate the agents. Develop the application logic, tools, knowledge retrieval, workflows, and connections to our business software. Establish reusable components so new agents can be added without rebuilding the entire system.

Make the systems reliable and manageable. Implement testing, monitoring, error reporting, retries, recovery procedures, and safeguards against duplicate or incorrectly completed actions. Provide a practical way for us to view agent activity, review exceptions, and pause or disable individual agents.

Protect access and data. Use secure remote access, appropriately limited permissions, protected credentials, and activity logs. Build approval checkpoints for actions such as publishing content, placing orders, changing important records, or sending sensitive communications.

Document, hand over, and improve the system. Maintain source code, configuration, deployment instructions, backups, and operating documentation in accounts we control. Train users, troubleshoot issues, and improve the agents based on measured results and business feedback.

Required Experience

• Demonstrated local AI deployment experience. You have personally deployed and operated language models and agents on physical hardware and can explain hardware requirements, memory usage, model selection, and performance tradeoffs.
• Strong software engineering skills. You can write, debug, test, and maintain custom application code, preferably in Python, with JavaScript or TypeScript experience where relevant.
• Agent and integration development. You have experience with tool-using agents, APIs, webhooks, authentication, databases, structured outputs, and document-based knowledge retrieval.
• Deployment and operations. You are comfortable with containers, service configuration, networking, monitoring, backups, and remotely maintaining locally hosted systems.
• Testing and security. You can demonstrate failure handling, access controls, output validation, and human approval mechanisms—not just successful demonstrations.
• Clear communication and ownership. You can translate business needs into technical requirements, explain tradeoffs, document your work, and independently move projects through implementation.

Experience limited to prompts, basic chatbot setup, or connecting cloud services through no-code tools is not sufficient by itself. Automation platforms are welcome where appropriate, but you must be able to develop and maintain custom code.

Particularly Relevant Experience

Experience with real-time voice agents, locally hosted speech recognition and speech generation, SEO and social media integrations, inventory systems, document extraction, and business-process automation will be valuable.

Experience with dental or healthcare workflows, practice-management software, and systems handling sensitive information is also a plus.

Expected Deliverables

The engagement should begin with a hardware assessment, proposed architecture, and prioritized development roadmap. From there, you will deliver a working pilot agent on our local hardware before expanding into additional agents.

Each completed agent must include its source code, deployment configuration, integration documentation, test results, operating instructions, and clearly documented limitations. The overall system should include monitoring, recovery procedures, and a maintainable structure for adding future agents.

Success will be measured by reliable task completion, accuracy, response times, reduced manual work, and ease of maintenance—not simply the number of agents created.

How to Apply

Please include:

1. Two or three relevant projects you personally built, explaining what each agent did, your individual contribution, and whether it was used in production.
2. Details of your local deployments: hardware specifications, models used, deployment approach, and any external services the system depended on.
3. Supporting evidence: a portfolio, code samples, recorded walkthrough, or a live demonstration that you are authorized to share.
4. Your proposed approach: how you would assess our machines, deploy the first agent, and support multiple agents over time.
5. Your availability, time zone, preferred working arrangement, and hourly or monthly compensation expectations.

Shortlisted candidates may be invited to complete a paid technical pilot in a restricted test environment before a broader engagement.

We are looking for someone who can take ownership of building a dependable, locally hosted AI system and continue developing agents that meaningfully improve how our business operates.