LINKEDIN ARTICLE — PIECE 5 OF 7
Publish: April 17, 2026 | Author: Rhyan J. Neble | ~1,100 words
The Small Business That Was Always Perfect for AI Agents (It's Not the One You Think)
Rhyan J. Neble | Founder & CEO, Extended Systems Intelligence | April 2026
Every AI vendor is pitching the same target customer right now: the mid-market enterprise. Healthcare systems. Financial services firms. Retail chains. Organizations large enough to have IT departments, large enough to absorb software costs, and large enough to appear in analyst market sizing reports.
Nobody is pitching rural ISPs. I want to make the case that this is a mistake — and that the Tier 2/3 ISP is actually closer to the ideal first customer for agentic AI than any of the obvious enterprise targets.
I spent years inside this market at ETI Software Solutions. Here is why it is structurally exceptional.
The ideal first market for agentic AI is one where the workflows are highly structured, the automation value is high, the data sovereignty requirements eliminate cloud alternatives, and the customer has an immediate, non-optional forcing function to act. That market is rural ISPs, today, in 2026.
Criterion 1: Highly Structured, High-Value Workflows
The most reliable agentic AI deployments operate on structured, repeatable workflows — the kind where the sequence of steps is well-defined, the inputs are machine-readable, and the correct output can be verified. Unstructured creative tasks, open-ended research, complex judgment calls — these are where agent error rates are highest and where human oversight remains essential.
ISP network operations are almost entirely structured workflows. An OLT fault has a specific set of diagnostic steps: check the upstream port, verify the optical levels, identify the affected ONTs, run the appropriate CLI sequence. The decision tree is not simple — it branches across dozens of device models, fault types, and network topologies — but it is structured. There is a correct answer. The agent can be evaluated against it.
FCC BDC compliance filing is even more structured. It is a data extraction, transformation, and submission task with a precisely defined schema, a defined submission window, and binary correctness — either the data is right or it isn't. An agent that handles BDC compliance filing is operating in one of the most favorable conditions for reliable autonomous performance.
This is not an accident. Network operations and regulatory compliance are domains that were designed for systematic execution. They were designed for humans who could follow complex procedures reliably — which means they were designed, whether intentionally or not, for agents.
Criterion 2: Data Sovereignty Is a Hard Requirement, Not a Preference
Cloud AI has a data sovereignty problem in regulated industries. For most markets, this is a preference — organizations would like to keep their data on-premises, but they're willing to accept cloud deployment with appropriate contractual protections.
For ISPs, it is not a preference. CPNI — Customer Proprietary Network Information — is a federal requirement that restricts disclosure and use of subscriber network data without consent. The network telemetry that drives fault diagnosis, the subscriber records that enable service management, the topology data underlying BDC compliance — this data cannot be routed through third-party cloud AI services without creating compliance exposure.
This eliminates the cloud AI vendors from a significant portion of the ISP workflow automation market. It does not eliminate on-premises agentic AI. In fact, it creates a structural competitive barrier around on-premises deployment that is regulatory in nature — much harder to overcome than the economic barriers that apply in non-regulated markets.
Criterion 3: BEAD Creates a Non-Optional Forcing Function
Most technology buyers have discretion about timing. They can evaluate AI tools, pilot them, delay adoption, revisit next quarter. The forcing function for adoption is rarely hard.
For ISPs receiving BEAD subgrant funding, it is hard. The Broadband Equity, Access, and Deployment programme distributes $42.45 billion in federal broadband infrastructure funding with the requirement that ISP subgrantees maintain accurate FCC BDC compliance documentation throughout their deployment periods. As network infrastructure expands under BEAD, the BDC compliance burden grows proportionally. An ISP that doubles its service area has a BDC filing that is roughly twice as complex.
NTIA has approved 53 of 56 state Final Proposals. Thirty-eight states have NIST clearance for fund access. Subgrantees are actively deploying capital now. The compliance clock is running. This is not a 'maybe next year' evaluation. It is an immediate operational need with a defined deadline and material consequences for non-compliance.
Criterion 4: The Beachhead Is Real and Bounded
One of the most common failures in enterprise software go-to-market is targeting a market that is too diffuse. The product is good for everyone and therefore hard to sell to anyone, because the sales motion requires educating each prospect from scratch about why this applies to them.
The Tier 2/3 ISP market is bounded and well-organized. WISPA has 3,000+ ISP members. They communicate through a defined set of industry conferences — Broadband Communities Summit, WISPA Annual, Mountain Connect. They share vendors — ETI Software Solutions serves a substantial portion of this market. They have shared vocabulary, shared regulatory obligations, and shared operational challenges.
A sale to one ISP operator creates a reference that is immediately legible to the next. They are at the same conferences, know the same people, face the same FCC compliance deadlines. This is word-of-mouth territory. The social proof from early deployments travels faster and lands harder than in a diffuse enterprise market.
Criterion 5: The Enterprise Vendors Are Structurally Absent
Microsoft Copilot, Salesforce Einstein, IBM watsonx — these platforms are not coming for the Tier 3 ISP market. The economics don't work: per-seat pricing at $30–$150/user/month is inaccessible for an ISP with a 12-person operations team and a $2M revenue base. The product doesn't fit: none of these platforms has Nokia, Calix, or Adtran CLI intelligence, GIS spatial network query capability, or FCC BDC compliance automation built in.
This is not a market where XSI Lodestone will face well-resourced direct competition in the near term. The enterprise AI vendors are serving enterprise customers. The ISP market is structurally too small for them and too specialized for a horizontal product to serve well. The competitive threat, if any, comes from a competitor who recognizes this market opportunity and moves quickly — not from Salesforce.
The Synthesis
The ideal first market for agentic AI is one where the workflows are highly structured (criterion 1), the data sovereignty requirements eliminate cloud alternatives (criterion 2), there is an immediate non-optional forcing function (criterion 3), the customer base is bounded and networked (criterion 4), and the enterprise vendors are absent (criterion 5).
The Tier 2/3 ISP market meets all five criteria simultaneously. That is not common. It is why this is the right first market, and why the window for establishing a first-mover position in it is measured in months rather than years.
Rhyan J. Neble | Founder & CEO, Extended Systems Intelligence | rneble@xtendedsystems.com | xsilodestone.ai
Q&A with Rhyan
Extended questions from discussions — answered in full.
Five criteria point to rural ISPs as ideal: (1) Highly structured, high-value workflows (fault diagnosis, BDC compliance), (2) Data sovereignty as a hard requirement eliminating cloud alternatives, (3) BEAD creating non-optional forcing function with imminent deadlines, (4) Bounded, well-organized customer base with tight communication networks, (5) Enterprise vendors structurally absent—Salesforce and Microsoft don't serve this market.
ISPs receiving BEAD subgrant funding must maintain accurate FCC BDC compliance documentation as their service areas expand. An ISP doubling its service area doubles its BDC filing complexity. NTIA has approved funding for 38 states with active capital deployment. The compliance clock is running with material consequences for non-compliance. This is not a 'maybe next year' evaluation but an immediate operational need.
The Tier 2/3 ISP market is bounded and well-organized: WISPA has 3,000+ members sharing vendors, vocabulary, and regulatory obligations. A sale to one operator creates legible social proof for the next. They are at the same conferences, know the same people, face the same deadlines. Word-of-mouth travels faster and lands harder than in diffuse enterprise markets.
ISP workflows are highly structured: OLT fault diagnosis has specific diagnostic steps, decision trees across device models, defined CLI sequences. These workflows were designed for humans who could follow complex procedures systematically—which means they were designed, whether intentionally or not, for agents. FCC BDC compliance is even more structured: data extraction, transformation, and submission with precisely defined schema and binary correctness.
Common Questions
Search-ready answers to the questions we hear most often.
Rural ISPs meet five criteria that make them ideal early adopters: (1) Highly structured, high-value workflows suited to agent automation, (2) Data sovereignty as hard regulatory requirement eliminating cloud alternatives, (3) BEAD creating non-optional forcing function with imminent deadlines, (4) Bounded, well-organized market with tight communication networks, (5) Enterprise vendors structurally absent from this market segment.
ISP workflows are highly structured: OLT fault diagnosis follows defined decision trees across device models with known CLI sequences; FCC BDC compliance is data extraction and transformation with precisely defined schema. These workflows were designed for humans following complex procedures systematically—which is exactly what agents excel at.
ISPs receiving BEAD subgrant funding must maintain accurate FCC BDC compliance through their deployment periods. As networks expand under BEAD, BDC complexity grows proportionally. NTIA has approved 38 states with active capital deployment. This is not a 'maybe next year' evaluation but an immediate operational need with defined deadlines and material compliance consequences.
WISPA has 3,000+ ISP members sharing vendors, vocabulary, and regulatory obligations. They communicate through defined conference circuits. A successful deployment creates legible social proof for peers—word-of-mouth travels faster and lands harder than in diffuse enterprise markets where each prospect requires custom education.
No. Salesforce, Microsoft, and IBM pricing ($30-150/user/month) is inaccessible for 12-person operation teams at $2M revenue ISPs. Their products lack Nokia/Calix/Adtran CLI intelligence, GIS spatial query capability, and FCC BDC compliance automation. Enterprise vendors serve enterprise customers. The rural ISP market is structurally too small and specialized.
The market opportunity is bounded and timing-sensitive. BEAD-funded ISPs have deployment windows and compliance deadlines measured in 18-36 months. The window for establishing first-mover advantage and building operational intelligence through deployments is measured in months, not years. Early movers establish reference relationships and deployment experience that later entrants cannot quickly replicate.