School System AI Software Adoption
Navigating Horizontal, Vertical, and Hybrid Solutions
Readiness Diagnostic
Before diving into the framework, take a moment to assess your organization's current state. These questions will help you identify opportunities and focus areas as you explore the guide.
When someone proposes adopting new AI software, how clear and repeatable is the decision-making process?
How aligned is your leadership team on when horizontal AI is sufficient versus when a vertical (or hybrid) solution is required?
How much visibility does your leadership team have into which AI software is currently being used across the organization or network?
Purpose and Overview
School systems today face a rapidly evolving AI landscape. To navigate it effectively, leaders need a clear, concise schema for understanding the market and underlying software architecture—because no single tool or category will meet all school system needs.
Most systems will ultimately require a portfolio of tools that together support a wide range of instructional and non-instructional school-based workflows.
This portfolio will likely include some combination of the following:
Horizontal (general-purpose) AI
Broad tools that support a wide range of operational and creative tasks
Hybrid AI (blended horizontal–vertical models)
Tools that combine general-purpose LLM capabilities with domain-specific scaffolds, guardrails, and curriculum alignment to deliver broader functionality while still meeting school-specific requirements
Vertical (narrow/domain-specific) AI
Purpose-built tools that align to curriculum, instructional models, or school system workflows and safety requirements
Key Concepts: Horizontal vs. Vertical vs. Hybrid AI
The categories below describe common design patterns in the AI software market. Individual tools may blur boundaries or evolve over time, and leaders should expect meaningful variation within each category.
3.1 Horizontal AI
"General-Purpose Engines"
Definition: General-purpose AI models and assistants trained on broad, internet-scale datasets and designed to work across domains and industries.
Illustrative Examples
Note: Several of these providers offer education-specific editions with additional guardrails, compliance features, and instructional scaffolds. These education editions often have hybrid characteristics and are discussed in Section 3.2.
Primary Value
- Versatile for many staff tasks (writing, summarizing, translating, coding)
- Rapid productivity gains with minimal configuration
- Often integrated into tools staff already use (email, documents, LMS, productivity suites)
Key Limitations for PreK–12
- Not inherently tuned to school system curriculum, standards, or pedagogical models by default, without significant customization or additional scaffolding
- May, by default, provide direct answers rather than supporting student thinking and mastery, particularly in student-facing contexts
- Requires strong usage guidance and data protections to avoid over-sharing student information
3.2 Hybrid AI
"Blended Horizontal–Vertical Models"
Definition: Hybrid AI tools combine the flexibility and generative power of horizontal models with the structured pedagogy, guardrails, and domain‑specific workflows of vertical applications.
Illustrative Examples
Hybrid tools exist on a spectrum. Some emphasize extensive purpose-built workflows; others emphasize model-level tuning with lighter workflow scaffolds.
MagicSchool — 80+ purpose-built educator tools and 50+ student tools, curriculum alignment via RAG, advanced moderation and safety features, usage dashboards. Sits toward the "workflow-heavy" end of the hybrid spectrum.
Playlab — Custom app-building platform where educators can create AI tools aligned to their curriculum, standards, and context. Emphasizes local customization and community sharing over pre-built workflows.
Gemini for Education / Google AI Pro for Education — Education-specific offerings that include model-level tuning (LearnLM integration optimized for learning), educator workflows (lesson planning, differentiation, assessment creation), distinct student experiences with additional guardrails for under-18 users, and custom Gems for education-specific assistants. Sits toward the "model-tuned with lighter scaffolds" end of the hybrid spectrum, with deep integration into Google Workspace.
ChatGPT Enterprise / ChatGPT Edu — Enterprise and education offerings that provide enhanced security, admin controls, and longer context windows. Custom GPTs allow creation of purpose-built assistants. These editions have some hybrid characteristics but fewer education-specific workflows than other options listed here.
Primary Value
- Broader functionality than traditional vertical tools while maintaining alignment to instructional models
- Often include built-in guardrails (such as prompt constraints, safe defaults, and, in some cases, audit logs) that can reduce the risk of misuse
- Can accelerate system-wide coherence when intentionally implemented and integrated into existing ecosystem workflows
Key Limitations
- Still require strict governance because underlying LLMs may behave like horizontal tools without proper constraints
- Risk of over-reliance if school systems mistake "hybrid" for fully compliant with all instructional or privacy requirements
- Quality varies widely depending on how much domain-specific structure the vendor has layered in
For many school systems, hybrid tools may become the "default middle path"—broad enough to be useful across roles, structured enough to be safe for instructional contexts, and more scalable than piecemeal vertical adoption.
3.3 Vertical AI
"Use Case Specific Applications"
Definition: Purpose-built AI applications designed for PreK–12 workflows—instructional, operational, talent, and compliance-related—often trained or configured using vetted curriculum, standards, or domain-specific business rules.
Illustrative Examples
Primary Value
- Purpose-built functionality: Vertical tools can build rich, nuanced workflows that horizontal models do not provide out-of-the-box
- Stronger alignment to school system curriculum and processes—whether instructional, operational, or talent-related
- Often include domain-specific guardrails: role-based access, audit trails, human-in-the-loop requirements, and compliance checks
- Designed to prompt, probe, and scaffold decisions or learning rather than simply generate answers
Key Limitations
- Narrower scope than horizontal models; each tool typically solves a specific workflow
- May create fragmentation if many single-purpose tools are adopted without integration
- Quality varies; "education-specific" branding does not guarantee strong pedagogy, strong privacy practices, or strong operational accuracy
- Even strong vertical tools depend on foundational model behavior beneath the surface, so governance and ongoing evaluation remain essential
3.4 Architecture Note
How Many AI Tools Are Actually Built
Most vertical and hybrid education tools do not rely on their own standalone AI models. Instead, they connect to large foundational models—such as those from OpenAI, Anthropic, or Google—through APIs (Application Programming Interfaces). An API is essentially a secure bridge that allows a tool to send a request to a foundational model (e.g., "rewrite this paragraph at a 5th‑grade level"), receive a response, and then apply additional instructional scaffolds, restrictions, guardrails, or workflows.
Because of this architecture:
- Vertical and hybrid tools often sit on top of horizontal foundational models.
- Some—but not all—vendors offer optionality, allowing school systems to choose which foundational model powers the tool.
- The tool's safety, alignment, and instructional quality depend on both the underlying model and the vendor's added structure.
This architectural layering is one reason the horizontal–vertical–hybrid landscape should be understood as a continuum, not rigid categories.
Check for Understanding
Now that you've learned about horizontal, vertical, and hybrid AI, test your understanding by categorizing these real-world scenarios.
Drag each scenario card to the category you think it belongs to.
An organization uses Google Workspace and has granted broad access to Gemini, which teams use across a wide range of workflows.
A school site adopts EnlightenAI, a purpose-built tool that provides students with actionable, rubric-aligned writing feedback.
A district adopts Playlab and grants broad access to staff. Playlab is designed for educators and enables users to create custom tools to support a variety of instructional and operational workflows.
Plot Your School System's Current Approach
Before exploring decision pathways, take a moment to reflect on where your school system currently falls on the AI portfolio spectrum.
Classify Your Horizontal AI Uses
Review how your school system currently uses horizontal AI tools across different domains. Rate each category's effectiveness and share insights on what's working or not.
If you have adopted vertical and/or hybrid solutions, what have the outcomes been?
Understanding where you currently are helps frame the decision pathways ahead. There is no single "right" position—the goal is to make deliberate, portfolio-level choices that match your system's needs, capacity, and strategic priorities.
Decision Pathways for AI Tool Choices
Once leaders understand the basic categories of horizontal, vertical, and hybrid tools, the next question is where a given use case should ultimately live. Before entering any pathway, leaders should be explicit about the problem they are trying to solve or the opportunity they are trying to capture—including who experiences it, what success would look like, and why current approaches are insufficient.
In practice, school systems often follow a small number of recurring decision pathways as they move from identifying a need to landing on a concrete tool or set of tools, including as AI-driven workflows and automations increase over time. Sometimes that journey ends in a vertical or hybrid product; sometimes it leads to an internally built solution; in other cases, the most durable answer is to keep the workflow inside a horizontal environment. This section names those patterns and surfaces key decisions leaders should revisit over time.
Pathway A: Start Horizontal → Stay Horizontal
When horizontal tools remain the best long-term home
In many cases, the most sustainable long-term home for a use case is a horizontal tool itself—especially when the work is already embedded in existing productivity suites.
Leaders determine that:
- horizontal tools are already tightly integrated into staff workflows
- the use case does not require complex workflows, permissions, or governance of student data
- the cost or overhead of adding another tool—or building and maintaining a custom solution—outweighs the benefits
The school system intentionally decides that the mature solution will remain in a horizontal environment (e.g., ChatGPT, Gemini, Copilot, Claude), supported by:
- clear guidance and exemplars for staff use
- training on privacy, equity, and verification
- proportionate monitoring of usage and impact
Pathway B: Start Horizontal → Build Internally
When custom internal solutions make the most sense
A rapidly expanding set of tools is lowering—but not eliminating—the barrier for educators, instructional coaches, and school system staff to build custom software without traditional development skills.
In some cases, instead of procuring a vertical or hybrid product, the school system builds a custom tool internally using approaches such as:
- Custom assistants within horizontal platforms (e.g., custom GPTs, Gems, Projects)
- Education-oriented builder platforms (e.g., Playlab)
- AI-assisted software development tools that allow natural-language app creation and workflow automation (e.g., Replit, Claude Code, Lovable, Bolt), enabling staff to generate and iterate on functional tools without traditional coding
Governance considerations include:
- Review and approval processes
- Documentation
- Monitoring for quality and safety
- Sustainability planning if the original builder moves on
- Clear boundaries around where internally built tools can be deployed (e.g., internal staff workflows versus student-facing use)
Illustrative examples: Internally built tools in this pathway might include a literacy coach's curriculum-aligned lesson-planning assistant, an operations intake and triage tool for central office requests, or a principal-facing observation feedback assistant aligned to the school system's instructional rubric.
Pathway C: Start Horizontal → Go Vertical
When purpose-built solutions are worth the investment
In other cases, the right answer is to move from horizontal experimentation into a purpose-built vertical or hybrid product.
This pathway is most appropriate when scaling requires:
- complex workflow automation with role-based permissions
- deep integration with SIS, LMS, or HRIS systems
- compliance features, audit trails, and monitoring at scale
- vendor-supported maintenance and updates
Pathway Permutations
Other valid entry points and variations
The pathways above are presented with horizontal tools as the most common starting point because this reflects how many school systems currently encounter and experiment with AI—often through existing productivity platforms. However, this is not the only valid entry point.
Some school systems begin with a vertical solution (e.g., a reading or tutoring tool tied to a specific instructional need), or adopt a hybrid platform as their first system-level investment.
These pathways are meant to illustrate common patterns, not prescribe a required sequence. Leaders should expect variation based on context, urgency, risk tolerance, and existing infrastructure.
Apply the Framework: Operations Help Ticket Scenario
Consider how you would apply the decision pathways to this real-world situation.
Scenario
School site operations teams report spending significant time tracking and following up on help tickets. While the team has streamlined intake through Google Form submissions, tickets realistically arrive through multiple channels—forms, email, hallway conversations, and ad hoc requests.
The team has leveraged their horizontal platform to improve this workflow with positive results. They've been able to review help ticket logs to triage requests and set up automated contact logs. While the system isn't perfect and doesn't handle the process end-to-end, it has demonstrably saved time and reduced manual entry for routine tasks.
Recently, an ops leader discovered a vertical AI platform specifically designed for help ticket management. It promises seamless integration of data from forms and email, intelligent prioritization, automatic assignment of due dates and task owners, and automated communications to all affected parties. However, the platform is expensive—potentially adding up to $10K annually depending on user accounts.
The team has also considered building custom software using AI coding agents, though they're uncertain whether it would match the sophistication of the vertical platform.
Using the framework's decision pathways, which approach would you recommend for this team?
Key Decisions Leaders Should Revisit Across Use Cases
Across these pathways, leaders can use a common set of questions to guide choices and revisit them over time as tools and needs evolve:
Should this live in a horizontal tool or in a vertical/hybrid product? (What are the privacy, workflow, and change-management implications of each?)
Should we build this internally using our horizontal tools? (Do we have the capacity to maintain it over time?)
Should we hire someone to build it for us using a horizontal model? (Is this important enough to warrant external expertise?)
Does an existing vertical or hybrid solution already meet most of our needs? (Are we better off configuring an existing product than maintaining our own build?)
If we go vertical or hybrid, should we commission or procure a custom solution? (Is the problem sufficiently unique or high-stakes to justify that investment?)
Used this way, the goal is not to "graduate" every successful use case into a vertical tool. Instead, the goal is to be explicit about why a use case is staying in a horizontal environment or moving into a vertical/hybrid product—and to revisit that decision as the system's strategy, risks, and opportunities evolve.
Implementation Recommendations
Section at a Glance
This section outlines three recommendations for strategic, scalable AI adoption in school systems:
- Establish a governed horizontal foundation so staff and students can use AI equitably and safely.
- Use disciplined pilots to evaluate when vertical or hybrid tools add real value beyond the core platform.
- Equip department leaders to set workflow expectations, ensuring quality, coherence, and human judgment.
Together, these recommendations are meant to be staged and iterative. School systems should expect to move back and forth between them as strategy, capacity, and tools evolve.
Recommendation 1: Adopt a Horizontal Enterprise Platform
A practical first move for most school systems is to adopt an enterprise-grade horizontal GenAI platform so that staff (and, where appropriate, students) can use powerful AI safely and equitably. Rather than starting with a patchwork of point solutions, an enterprise horizontal platform creates a common foundation for experimentation, capacity building, and early wins across instructional, operational, talent, and communications workflows.
As a starting point, most school systems should first enable the enterprise AI that comes with their existing productivity suite (e.g., Google Workspace or Microsoft 365). This minimizes additional procurement, reduces change-management friction, and allows leaders to focus on norms, guidance, and early use cases rather than standing up a brand‑new platform.
A note on education editions: The major platform providers increasingly offer education-specific versions of their AI tools—such as Gemini for Education, Google AI Pro for Education, and ChatGPT Edu—that include compliance features, admin controls, and varying degrees of instructional scaffolding. These education editions often have hybrid characteristics (see Section 3.2), which means that adopting an "enterprise horizontal platform" may also provide some of the benefits traditionally associated with vertical or hybrid tools. Leaders should evaluate what their existing productivity suite's education AI offering includes before assuming they need to procure additional hybrid or vertical solutions for basic use cases.
Comparison: Without vs. With a Horizontal Enterprise AI Platform
| Dimension | Without a Horizontal Enterprise AI Platform | With a Horizontal Enterprise AI Platform |
|---|---|---|
| Equity in Access to Top GenAI Tech | Many staff use free GenAI tools with limited features and weaker protections. A small number of staff purchase their own access to stronger GenAI tools, creating uneven access. Students often have no sanctioned access at all. | All staff (and, where appropriate, students) can be given equitable access to high-performing GenAI tools under a single, governed environment. Access levels can be differentiated by role and age, rather than by who can pay or experiment on their own. Leaders can communicate a clear "green list" of approved tools and entry points. |
| Equity in Ability to Build and Share AI Tools | Many leaders, staff, and students are capable of building simple AI-powered tools, but only a small subset have access to the GenAI accounts that make this possible. Even when useful tools are built, there is no easy way to share them across schools or teams. Innovation is fragmented and often invisible to system leaders. | Networks can intentionally build a culture in which leaders, staff, and students become builders of AI tools that improve key workflows. Reusable prompts, templates, and custom assistants can be created once and shared widely across the organization. High-leverage tools can be curated, improved, and incorporated into training and onboarding. |
| Admin Visibility & Insight | Most GenAI usage happens in individual, consumer accounts outside district oversight. Leaders have little visibility into who is using GenAI, for what purposes, or with what outcomes. It is difficult to distinguish promising innovation from risky or duplicative usage. | Organizational leaders can view aggregate GenAI usage patterns, such as adoption by role, common use cases, and growth over time. Leaders can see which internally built tools are being used and by whom, informing investment and support decisions. Data can inform professional learning, guardrail updates, and future procurement. |
| Data Privacy & Protection | Staff may unknowingly input PII or sensitive information into tools with weak or unclear data protections. There is no consistent way to enforce district privacy expectations across individual accounts. Leaders face elevated risk without corresponding insight or control. | Usage can be governed under a single data protection agreement and clear terms of use. Technical controls and training can reduce the likelihood that PII or sensitive data are shared inappropriately. Leaders can align GenAI use with existing privacy, security, and records policies, and adjust settings as policies evolve. |
Over time, this horizontal enterprise layer becomes the base of the AI stack: a shared environment for experimentation, everyday productivity, and early tool-building that complements, rather than replaces, targeted vertical and hybrid solutions.
Recommendation 2: Conduct Small-Scale Pilots of Vertical and Hybrid Solutions
While an enterprise horizontal platform provides a common foundation, most school systems will still need targeted hybrid and vertical tools for specific instructional, operational, talent, or compliance workflows. Rather than jumping directly to large procurements, leaders should run small, time-bound pilots led by domain experts.
What Matters Most in Pilots
- Pilot impact: Did the pilot demonstrate clear, measurable benefits aligned with the initial vision of success for students, staff, or system workflows?
- Equity monitoring: Were outcomes examined by student subgroup (e.g., race/ethnicity, disability status, English learner status, grade level)? Did any subgroups experience neutral or negative outcomes even if aggregate results were positive? What unintended consequences emerged, and for whom?
- Pedagogical integration: How did the tool change teaching practice? Did it uphold or collapse rigor? Did it align with the school system's instructional model and pedagogical priorities? What tradeoffs emerged between efficiency gains and instructional quality?
- Scaling implications (data and IT): Have the data and IT implications of expanded deployment been considered (e.g., integrations, security posture, data flows, support model)?
- Scaling implications (staffing and professional learning): Have the staffing, training, and professional learning implications been considered (e.g., who supports, maintains, and champions the tool; time, content, and ownership for PD)?
- Roles and ownership: Is there a clear long-term owner for this initiative beyond the pilot phase, at both the system and school levels where relevant?
- Budget and sustainability: Is there a viable long-term funding source or budget allocation for scaling, and does it align with broader portfolio and vendor-management strategy?
Ending a Pilot is a Success Condition
Not every pilot should scale—and that is not a failure. Leaders should consider ending a pilot when the cost–benefit no longer warrants continuation, when it fails to address the original problem, creates unintended harm, requires disproportionate support, or is surpassed by a better alternative. Importantly, pilots should be treated as learning experiences: it is healthy—and expected—for some pilots not to succeed. In fact, if a school system has no pilots that end, it may be a signal that it is not testing enough new ideas.
Recommendation 3: Prepare Department Leaders to Set AI Workflow Expectations
Whether a workflow lives in a horizontal, hybrid, or vertical tool, the introduction of AI creates quality-control risks. AI can mis-handle quantitative analysis, produce weak or misaligned lesson plans, generate inaccurate summaries, or drift away from the standards leaders expect in curriculum, communications, or operational documents.
Department leaders (instruction, special education, HR and talent, operations, finance, communications) are best positioned to:
- Identify the critical workflows in their domain and clarify where AI is expected, optional, or not appropriate
- Name specific quality risks (e.g., math errors, misaligned scope and sequence, inappropriate tone in family communications) and set explicit human-review expectations
- Provide concrete examples and non-examples of acceptable AI use in their domain, including which portions of a workflow AI can support versus which must remain human-led
Implicit in this recommendation is that department leaders must be active users of AI tools themselves. When leaders do not set clear expectations, AI use tends to fragment into inconsistent quality, shadow workflows, and uneven risk exposure across schools. Leaders who regularly test horizontal, hybrid, and vertical solutions in their own work will be far better equipped to:
- Offer nuanced, credible guidance to staff
- Spot emerging risks before they scale
- Continuously refine workflow expectations as tools and use cases evolve
Case Studies
Explore how Breaking Ground Prep—a four-school network—approached AI adoption decisions across talent, academics, and operations.
Talent Team: Streamlining Interview Processes
Breaking Ground Prep · Four-school network, ~30 hires/year
Breaking Ground Prep is a four-school network that hires approximately 30 new staff members each year. The organization uses an organization-wide, general-purpose AI platform, and all staff have been trained on it. This platform functions as a horizontal AI tool across the organization.
Value from Horizontal AI
Within the talent team, the horizontal AI platform has been particularly valuable for refining job descriptions and drafting interview tasks and scorecards. These uses have streamlined interview preparation and led to more consistent, higher-quality hiring materials and processes. While interview scheduling remains somewhat manual, all hiring managers maintain Calendly availability, and sharing links with candidates has proven sufficient.
Evaluating a Vertical Solution
Recently, a member of the talent team began exploring vertical AI software designed specifically for HR and talent workflows. Many of these platforms advertise end-to-end capabilities, including interview and scorecard generation, automated scheduling, and even AI-conducted first-round interviews. Impressed by the promise of further efficiency gains, the team member gathered several vendor quotes and brought a proposal to the Senior Director of Talent.
The Decision
After careful consideration, the Senior Director of Talent decided not to pilot a new tool. The team was already realizing significant value from their horizontal AI platform for interview preparation, reducing the marginal benefit of adopting a specialized vertical platform. While automated scheduling might improve convenience, Calendly addressed the core need adequately. Most importantly, as a relatively small network with a limited number of annual hires, replacing a human phone screen with an AI-led interview risked weakening an important early candidate touchpoint—an unacceptable tradeoff given the modest time savings.
What does this case suggest about when a horizontal AI tool is sufficient—and when a vertical solution is worth the additional complexity?
SavedAcademics Team: Improving Writing Feedback
Breaking Ground Prep · Four-school network, middle and high school
Breaking Ground Prep is a four-school network serving middle and high school students. The organization uses an organization-wide, general-purpose AI platform, and all staff have been trained on it. This platform functions as a horizontal AI tool across the network. Teachers regularly use the platform as a collaborative partner for lesson design and instructional planning, finding it helpful for brainstorming and preparation.
Strategic Priority: Student Writing
This year, Breaking Ground placed a strategic emphasis on improving student writing. Leadership identified high-quality, timely feedback on student writing as a key instructional lever for improving outcomes. While teachers attempted to use their horizontal AI platform to support writing feedback, these efforts produced mixed results.
Piloting a Vertical Solution
During this period, one principal learned about a vertical AI platform designed specifically to provide instant, rubric-aligned feedback on student writing. Seeing a potential fit with the network's instructional priorities, the principal proposed a pilot to the Chief Academic Officer (CAO).
The CAO and principal selected four teachers to participate in a structured pilot of the vertical solution. The pilot group met regularly to share lessons learned, compare instructional practices, and refine how the tool was used with students.
Pilot Outcomes
After three months, the team identified several clear outcomes:
- Teachers saved significant time providing writing feedback
- Student feedback was consistently strong and closely aligned to rubrics
- Teachers observed measurable improvements in student writing, particularly in students' ability to link evidence to claims
Based on these results, the leadership team decided to expand access to the vertical platform to all humanities teachers at the pilot school. The network also set a goal of rolling the program out across all schools the following fall, leveraging the pilot teachers to lead training grounded in real classroom experience.
What does this case suggest about when a horizontal AI tool is sufficient—and when a vertical solution is worth the additional complexity?
SavedOperations: Transportation Logistics
Breaking Ground Prep · Four-school network, middle and high school
Breaking Ground Prep is a four-school network serving middle and high school students. The organization has adopted a general-purpose, horizontal AI platform that all staff are trained to use across functions. The operations team regularly relies on this tool for a range of workflows and often finds the output helpful for planning and documentation.
The Transportation Challenge
One persistent operational challenge, however, has been student transportation. Each year, planning bus routes that balance pickup locations, commute times, and operational constraints requires weeks of manual work. Despite repeated attempts, the team was unable to improve this process using their horizontal AI platform.
Exploring a Vertical Solution
Recently, a school-level Director of Operations learned about a vertical AI solution purpose-built for transportation. The platform automatically generates bus routes based on addresses, tracks buses in real time, and proactively communicates with families when buses are delayed, including updated estimated arrival times. Intrigued by the potential impact, the network approved a pilot at a single school site.
Pilot Results
Early results were promising:
- Route creation, which previously took weeks, was completed in a matter of hours, allowing the team to focus on reviewing and refining routes rather than building them from scratch
- One month into the school year, the site recorded only ten family complaints related to late buses—a noticeable improvement based on past experience
- Site operations leaders reported a significant reduction in stress, as they no longer needed to manually monitor buses or send ad hoc updates to families throughout the day
Scaling the Solution
Based on these outcomes, the network decided to expand the transportation platform across all schools the following year. The rollout plan included structured training and shared best practices developed by the pilot site, ensuring that lessons learned translated into smoother implementation network-wide.
What does this case suggest about when a horizontal AI tool is sufficient—and when a vertical solution is worth the additional complexity?
SavedClosing
A coherent AI strategy will not emerge from chasing tools, but from leaders making deliberate, portfolio-level choices about when to rely on horizontal platforms, when to invest in vertical or hybrid solutions, and when to say no.
By using this schema, starting with an enterprise horizontal foundation, and running disciplined, expert-led pilots, school systems can harness AI to strengthen instruction, operations, and talent—without losing sight of equity, safety, or human judgment.