Situation: A 150-person healthcare services company provides 24/7 remote behavioral monitoring and support for developmentally disabled adults living independently. Each individual they support has unique behavioral patterns, safety needs, and daily routines, requiring highly tailored monitoring and rapid, accurate incident detection. Their operations run continuously and must comply with strict state and county regulatory requirements for reporting, documentation, and appropriate use of technology.

The company is already highly innovative, with deep experience using IoT sensors and connected technologies to deliver personalized support at scale. Now, leadership wants to take the next step: apply AI to reduce manual workload, improve accuracy, strengthen compliance, and increase profitability.
But even with their vision and technical maturity, the rapid pace of AI developments — together with limited internal bandwidth and specialized AI expertise — made it difficult to evaluate options, avoid missteps, and confidently prioritize the right initiatives. They needed ongoing AI leadership to guide strategy, build capability, and help the company stay ahead of the curve.

They partnered with SoT to serve as their fractional Chief AI Officer team, embedded directly into leadership and operational teams.

What we are doing (ongoing): SoT stepped in as the company’s fractional CAIO, providing AI leadership, strategic direction, and hands-on execution across the business. Our primary focus is building a sustainable, scalable AI capability that the organization can rely on for years to come, as well as to support the immediate execution of ongoing AI projects.

Working closely with the COO and department leaders across Remote Support, Client Care, Field Services, Billing, Sales, Marketing, and Engineering, we:

• Establish the foundation for an enterprise AI capability, including programs, processes, and decision frameworks to operationalize AI in a structured, repeatable way.
• Introduced a formal AI portfolio management process to evaluate, compare, and prioritize new and existing AI opportunities based on impact, feasibility, risk, and regulatory considerations.
• Designed a rapid intake workflow for surfacing AI ideas from across the company and external partners, allowing promising opportunities to be vetted quickly and objectively before entering the portfolio.
• Created a vendor and supplier intake process to systematically evaluate AI tools and proposals and align them with internal needs and portfolio priorities.
• Built a market and technology intelligence capability to track emerging AI models, tools, and solutions — including early-stage but potentially disruptive technologies — so the company can time adoption and prepare ahead of the curve.
• Reviewed and strengthened the existing AI project portfolio, adding structure, clarity, and alignment to business goals.
• Began shaping AI governance, including early considerations around safety, compliance, appropriate use, and integration with regulatory obligations.
• Supported ongoing tactical AI initiatives, providing hands-on leadership and guidance to ensure teams can progress confidently and effectively.

Early Outcomes

Even at this early stage, the company is experiencing strong momentum and clarity:

• A unified, structured AI portfolio spanning all major business functions
• Clear prioritization of high-value opportunities and elimination of unrealistic or low-impact ideas
• A consistent intake & prioritization process, preventing wasted effort and ensuring alignment
• Stronger cross-department alignment on goals and where AI can make the biggest difference
• Early foundations for AI governance and capability building
• Executive-level enthusiasm and commitment, increasing internal support for AI initiatives

Upcoming milestones include finalizing the prioritized portfolio, establishing the AI roadmap, and identifying candidates for early pilots.

 

Go to the Top

Situation: A commercial Unmanned Aerial Vehicle (UAV)/drone operator wanted to expand beyond flight services and uncover new ways to create value using the large volumes of data they collect. They believed AI, especially generative AI, could unlock new insights and automation across the entire drone operations lifecycle, from pre-flight planning and regulatory preparation to post-flight data analysis and reporting.

The company partnered with SoT to identify a high-impact starting point, test feasibility quickly, and determine whether a focused AI capability could become a scalable service, or even a new business line.

What We Did (Ongoing)

1. Deep dive into a high-impact opportunity. Working with the operator’s leadership, we jointly selected one one specific use case of using generative AI to help verify FAA regulatory compliance during preflight planning. Together, we:

• Defined the scope and boundaries of the compliance challenge
• Clarified what “good enough” looks like for operators and regulators
• Identified the relevant data sources (FAA, 3rd party, and other agencies)
• Documented requirements and constraints for a realistic proof of concept

2. Designing and building the PoC on SoT’s innovation platform. With the scope and requirements understood, SoT:

• Designed a proof of concept specifically for preflight regulatory compliance checks
• Implemented the PoC on our cloud-based generative AI innovation platform, using generative AI to analyze mission details against relevant FAA guidance and restrictions
• Deployed the PoC for testing with the operator’s team

From start to finish, the process was completed in two months. Testing showed that the PoC was technically feasible and could meaningfully support preflight regulatory compliance activities.

3. Evaluating SaaS potential and business feasibility. After the successful PoC, the operator wanted to understand whether this capability could become a commercially viable SaaS service for other commercial drone operators. SoT worked with them to:

• Define an initial set of functional requirements and features for a potential SaaS offering
• Estimate the costs to build and operate the service
• Develop financial and revenue projections for different adoption and pricing scenarios
• Assess overall feasibility and potential business impact

The initial analysis showed promising business potential.

4. Moving into market validation and investment planning (in progress). With a successful PoC and a promising business opportunity, our focus has shifted to validation and funding. We are now helping the operator:

• Design and launch a market and customer validation effort to understand demand, priority features, and willingness to pay
• Translate validation results into beta product requirements
• Re-estimate build and operational costs based on refined scope
• Develop a business plan and investment strategy to support fundraising for product build-out

Results/Early Outcomes

• Successful PoC demonstrating generative AI can assist with FAA compliance checks in preflight planning
• Clear, scoped SaaS concept with defined functional requirements
• Cost and revenue models to evaluate business feasibility
• Promising financial outlook, justifying further market validation
• A structured path from PoC → beta product → potential new business line

 

Go to the Top

Situation: As part of our internal SoT AI initiative, Strategy of Things is actively transforming itself into an AI-native company. That means going beyond using AI as a productivity tool and instead building the operational infrastructure, workflows, and governance needed to run AI agents as part of a real Digital Workforce.

To do that, we launched Beacon, our internal AI Chief of Staff, built on OpenClaw, an open-source AI agent infrastructure platform. The goal was not to create a demo or experiment in isolation, but to learn firsthand what it actually takes to design, deploy, operate, and govern an AI agent that supports real day-to-day business activities.

We wanted to better understand questions that many organizations are now beginning to face: What does a production-grade AI agent require to operate reliably? What kind of memory, retrieval, automation, governance, and monitoring are needed? And how do you structure AI agents so they are trustworthy, scalable, and useful in real business settings?

By building and operating Beacon ourselves, we are using our own business as a live testbed for AI-native operations and Digital Workforce design.

What We Did (Ongoing): SoT designed and deployed Beacon as an internal AI Chief of Staff that supports our advisory and research operations. Built on OpenClaw and integrated with a structured knowledge base and operational workflows, Beacon is being used to help us validate how AI agents can function as persistent, role-based digital workers inside a business environment.

Our ongoing work includes:

• Designing and operating Beacon as a role-specific institutional agent for SoT, focused on research, knowledge management, operational continuity, and internal support
• Building and maintaining a structured knowledge base to give Beacon persistent institutional memory across frameworks, documents, decisions, and research assets
• Using semantic and full-text retrieval to help Beacon search and synthesize relevant internal context
• Running Beacon through an asynchronous operating model so it can support work across time zones and workflows without losing context
• Implementing automations, delivery workflows, retry logic, and health checks that help move the system beyond chatbot behavior into production-style agent operations
• Testing a two-agent architecture that separates organizational context from personal context, helping us validate the importance of role-scoped AI agents with clear boundaries
• Evaluating how model selection, memory design, governance controls, and observability affect agent reliability and usefulness in real-world use
• Documenting lessons learned so they can inform how we advise clients on agent platforms, Digital Workforce design, AI governance, and AI-native operating models

Results/Early Outcomes: Even at this stage, the Beacon initiative is already producing important operational and strategic value for SoT:

• A functioning internal AI Chief of Staff that supports knowledge retrieval, research synthesis, and operational continuity
• A real-world proving ground for understanding what is required to run AI agents reliably in day-to-day business operations
• Practical validation that memory architecture, operational workflows, and governance matter as much as model capability
• Early evidence that role-scoped agents with clear responsibilities are more governable, trustworthy, and effective than monolithic assistants
• A stronger internal foundation for SoT’s own AI-native transformation and Digital Workforce strategy
• Direct experience with the operational realities of agent deployment, including monitoring, recovery procedures, model fit, and production readiness
• Valuable lessons we can apply to client work involving AI agents, AI operating models, Digital Workforce design, and enterprise AI governance

 

Go to the Top