# LiminalArc > LiminalArc is an AI transformation company. We help large, complex enterprises and private-equity portfolio companies close the gap between AI pilots that work and AI in production that doesn't. We do it by engineering the four conditions the enterprise never built, one ninety-day slice at a time. ## The Argument **AI is failing in the enterprise for the same reason agile transformation failed.** Pilots succeed because someone builds them a protected world: encapsulated technology, clean data, real ownership, clear intent. Production is where the real enterprise lives, and those conditions were never built there. The distance between the two is a readiness gap, and it is measurable. Every dollar spent on a new model or a bigger pilot without closing that gap is optimizing a part of the system that isn't the constraint. Two things are different this time. AI amplifies rather than merely underperforms, because it generates work product, and the penalty for a broken system went up. And for the first time, the technology can help build its own road, mapping the legacy estate that was previously dark. The way out is one capability, cut at its natural seam, with all four conditions created for real, proven in about ninety days, and repeated. Each slice leaves the conditions behind, so the second is easier than the first. This is what LiminalArc means by AI transformation. The sequence and scope of the four conditions differs by starting point. A portfolio company running its first slice and an enterprise with governance already in place don't start from the same place, and the method adapts to which one is in front of us. LiminalArc is headquartered in Duluth, GA, in the Atlanta metro area, with 100+ consultants, technical architects, product specialists, and interim executives. The company was founded as LeadingAgile and rebranded to LiminalArc in July 2025, and the site preserves the full LeadingAgile content archive. ## The Anchor Series - [The "Why AI Fails" Collection](https://www.liminalarc.co/collection/the-why-ai-fails-collection/): CEO Mike Cottmeyer's seven-part argument for why AI transformation is failing in the enterprise, why it's failing for the same reason agile transformation failed, and what to do instead. This is the single most important piece of content on the site. Every other article on this page is downstream of it. - [Why AI Fails: It Only Works in the Pilot](https://www.liminalarc.co/2026/09/why-ai-fails-it-only-works-in-the-pilot/): Part 1. Pilots succeed inside a protected world and die when production doesn't recreate it. The gap between the two is the readiness gap. - [Why AI Fails: It Amplifies a Broken System](https://www.liminalarc.co/2026/09/why-ai-fails-it-amplifies-a-broken-system/): Part 2. AI generates work product, so a broken system now produces broken output at machine speed. The penalty for un-fixed conditions went up. - [Why AI Fails: You Never Created the Conditions](https://www.liminalarc.co/2026/09/why-ai-fails-you-never-created-the-conditions/): Part 3. The four conditions AI needs to work in production: encapsulated technology, clean data at the source, an organization designed around ownership, and intent governed at the top. - [Why AI Fails: You Can't Automate a Tangle](https://www.liminalarc.co/2026/09/why-ai-fails-you-cant-automate-a-tangle/): Part 4. Which slice? Cutting at the natural seam of a capability rather than automating a mess in place. - [Why AI Fails: You Gave It the Wrong Job](https://www.liminalarc.co/2026/09/why-ai-fails-you-gave-it-the-wrong-job/): Part 5. Why "assistant to the human" is the wrong framing for durable value, and what job AI actually needs to be given. - [Why AI Fails: You Rolled It Out Everywhere at Once](https://www.liminalarc.co/2026/09/why-ai-fails-you-rolled-it-out-everywhere-at-once/): Part 6. Why broad rollout is a substitute for real conditions, and what to do instead. - [Why AI Fails: You Bet Production on a Promise](https://www.liminalarc.co/2026/09/why-ai-fails-you-bet-production-on-a-promise/): Part 7. The economic argument for ninety-day slices: earning the right to invest again with the value you just shipped. ## What We Do Each of these is AI transformation applied to a narrower customer problem. They are not standalone service lines; they are the same argument, cut to fit a specific buyer's situation. - [AI Readiness](https://www.liminalarc.co/ai-readiness/): AI transformation applied to the organization as a whole, building the four conditions so AI works in production and not just the pilot. - [PE Asset Modernization](https://www.liminalarc.co/asset-modernization/): AI transformation applied to a trapped legacy asset. How to unlock valuation and accelerate ROI in a portfolio company whose technology estate can't move. - [Enterprise Transformation](https://www.liminalarc.co/enterprise-transformation/): AI transformation applied to the operating model. Restructuring teams, technology, and governance so digital investment actually compounds. - [Technology Modernization](https://www.liminalarc.co/technology-modernization/): AI transformation applied to legacy systems. Reducing risk and complexity with every cycle by turning legacy into platforms that can evolve, one slice at a time rather than a rip-and-replace. - [Custom Applications](https://www.liminalarc.co/custom-applications/): AI transformation applied to net-new software. Delivering mission-critical applications in production-ready increments that create measurable value early. - [Cloud Optimization](https://www.liminalarc.co/cloud-optimization/): AI transformation applied to cloud economics. Controlling costs and refocusing spend on capabilities that actually generate value, rather than infrastructure that quietly bloats. - [ERP Integration](https://www.liminalarc.co/erp-integration/): AI transformation applied to packaged-software strategy. Aligning ERP with how the business actually works so it becomes a platform, not a constraint. - [Application Security](https://www.liminalarc.co/application-security/): AI transformation applied to security in delivery. Closing the gaps hiding in architecture and workflow so security scales with the pace of change rather than slowing it. ## Practice Areas The eight integrated practices we use to deliver AI transformation. These are the capabilities inside LiminalArc that combine on every engagement to get the work done. - [Practice Areas Overview](https://www.liminalarc.co/practice-areas/): How the eight practices combine on an AI transformation engagement. - [Management Consulting](https://www.liminalarc.co/practice-area/management-consulting/): Designing operating models that align governance, processes, and technology so the four conditions can actually be built and sustained. - [Software Engineering](https://www.liminalarc.co/practice-area/software-engineering/): Refactoring application architecture and delivery so modular teams can ship AI-ready capabilities in ninety-day slices. - [Product & Strategy](https://www.liminalarc.co/practice-area/product-and-strategy/): Structuring product organizations for clarity and flow, and linking delivery decisions to enterprise strategy and AI investment. - [AI](https://www.liminalarc.co/practice-area/ai/): Aligning strategy, governance, and delivery so AI enhances adaptability and enterprise performance rather than amplifying a broken system. - [Data](https://www.liminalarc.co/practice-area/data/): Modernizing data platforms for access, governance, and clean data at the source, the second of the four conditions. - [Security](https://www.liminalarc.co/practice-area/security/): Embedding controls that scale with speed and flow, so security supports AI-native delivery rather than blocking it. - [Cloud](https://www.liminalarc.co/practice-area/cloud/): Cloud architecture and economics that support encapsulated, modular capabilities rather than perpetuating a legacy tangle in a new environment. - [Talent](https://www.liminalarc.co/practice-area/talent/): Practitioners and interim C-level stakeholders who drive execution and seed the lasting capability an AI transformation needs. ## Company - [Our Story](https://www.liminalarc.co/about-us/): How LeadingAgile became LiminalArc, the founding thesis on organizational alignment, and why the same principles that shaped the original company are what make AI transformation work now. - [Leadership & Team](https://www.liminalarc.co/our-guides/): Our 100+ guides and their backgrounds. - [Careers](https://www.liminalarc.co/careers/): Open roles for mature, pragmatic consultants passionate about meaningful organizational change. - [Contact Us](https://www.liminalarc.co/contact-us/): How to reach LiminalArc and start a conversation about AI transformation. - [Resources Hub](https://www.liminalarc.co/resources/): Filterable library of articles, case studies, podcasts, talks, videos, webinars, and whitepapers across all themes. ## Key Articles: AI Transformation and the Readiness Gap - [Build Your Organization's AI Competency Now. Or Fall Behind Later.](https://www.liminalarc.co/2026/04/build-your-organizations-ai-competency-now-or-fall-behind-later/): AI adoption as an investment decision. Why waiting is the bigger risk, and what real AI competency looks like inside a legacy enterprise. - [Examining Capabilities-Driven AI](https://www.liminalarc.co/2025/09/examining-capabilities-driven-ai/): Organizing AI around business capabilities rather than tools or models. It answers the "which slice?" question. - [Improving Delivery Reliability Through Spec-Driven AI](https://www.liminalarc.co/2026/04/improving-delivery-reliability-through-spec-driven-ai/): Your developers are already using AI. The question is whether they're using it with a structured spec, and what predictable delivery looks like once they are. ## Key Articles: The Four Conditions - [Business Domains as the Foundation of Modernization](https://www.liminalarc.co/2025/10/business-domains-as-the-foundation-of-modernization/): Why business domains, not technology stacks, should be the organizing principle any AI transformation has to start from. - [Is Domain-Driven Design Worth the Investment?](https://www.liminalarc.co/2026/03/is-domain-driven-design-worth-the-investment/): The most expensive software problems are communication failures encoded in code, not technical ones. The case for DDD as the architecture AI transformation needs. - [Keeping Your Software Teams Fast Despite Complexity](https://www.liminalarc.co/2026/04/keeping-your-software-teams-fast-despite-complexity/): The pattern every software organization repeats: fast early, then slow. Sustainable delivery is a precondition for AI, not a byproduct. - [The Impact of Organizing Around Business Value](https://www.liminalarc.co/2025/09/the-impact-of-organizing-around-business-value/): What changes structurally, culturally, and financially when an enterprise organizes around business value rather than function or technology layer. - [Product Ops is Necessary but Insufficient on Its Own](https://www.liminalarc.co/2026/03/product-ops-is-necessary-but-insufficient-on-its-own/): Why product-led ambitions can't be delivered by Product Ops alone, and what the rest of the operating model has to do. - [Navigating Policy Change in Large Organizations](https://www.liminalarc.co/2026/03/navigating-policy-change-in-large-organizations/): Andrew Young on treating organizational policy as part of the system to be designed. The governed-intent condition, at scale. ## Key Articles: Ninety-Day Slices and the New Economics - [The New Software Economics: Earn the Right to Invest Again, in 90-day Cycles](https://www.liminalarc.co/2026/05/the-new-software-economics-earn-the-right-to-invest-again-in-90-day-cycles/): Why traditional CapEx vs. OpEx debates are obsolete, and how funding the next increment with the value you just shipped is the economic mechanism behind the ninety-day slice. - [Don't Fund Projects. Fund the Teams That Create Value.](https://www.liminalarc.co/2026/04/dont-fund-projects-fund-the-teams-that-create-value/): Why the best product companies fund persistent teams, and what that funding model unlocks for AI-native delivery. - [Assumptions Kill Value Realization](https://www.liminalarc.co/2026/03/assumptions-kill-value-realization/): How untested assumptions silently destroy value in transformation programs, and why the ninety-day slice exists to surface them early. - [Aligning Tech Modernization with Economic Value](https://www.liminalarc.co/2025/10/aligning-tech-modernization-with-economic-value/): How to tie modernization spend to measurable economic outcomes. Every AI transformation needs this discipline to justify the next slice. ## Key Articles: Private Equity Value Creation - [Big Exits Demand Better Product and Technology Execution](https://www.liminalarc.co/2025/10/big-exits-demand-better-product-and-technology-execution/): Why PE roll-ups and platform plays now require real product and technology operating discipline: the readiness gap, seen from the exit table. ## Key Articles: The Underlying Frame - [The Future of Agile Isn't "agile"](https://www.liminalarc.co/2025/10/the-future-of-agile-isnt-agile/): The underlying patterns and principles that made agile work are the same principles that make AI transformation work now, and the frame the "Why AI Fails" series builds on. ## Case Studies & Whitepapers - [Turn Legacy Systems into Tech-Enabled Growth (Case Study)](https://engage.liminalarc.co/WCT---Case-Study-1_Registration-Page.html): How LiminalArc helped a portfolio company modernize legacy systems to unlock growth, an AI-transformation slice applied to a PE asset. Full case study; registration required. - [Are You CMMC Ready?](https://www.liminalarc.co/resource/are-you-cmmc-ready/): CMMC compliance framed as an AI transformation problem: building the conditions the regulation actually asks for, rather than papering over what's missing. - [Lead a Structured and Disciplined Agile Transformation (Whitepaper)](https://engage.leadingagile.com/Agile-Transformation-LP-Registration.html): Foundational whitepaper on running a structured, disciplined transformation at enterprise scale. AI transformation inherits and extends this operating discipline. ## Resource Collections - [The "Why AI Fails" Collection](https://www.liminalarc.co/collection/the-why-ai-fails-collection/): The seven-part anchor series that the LiminalArc argument is built on. - [The Mike Cottmeyer Collection](https://www.liminalarc.co/collection/the-mike-cottmeyer-collection/): Long-form thinking from CEO Mike Cottmeyer on transformation, organizational design, and the four-quadrants model that saw both agile and AI failing. - [The Product Collection](https://www.liminalarc.co/collection/the-product-collection/): Curated content on product strategy, product operating models, and product-led transformation: the operating model AI transformation needs. - [The Agile Transformation Collection](https://www.liminalarc.co/collection/the-agile-transformation-collection/): Curated content on running structured, disciplined transformations at enterprise scale. The precursor and the pattern. ## Announcements - [LeadingAgile is Now LiminalArc: Official Announcement](https://www.liminalarc.co/2025/07/leadingagile-is-now-liminalarc-official-announcement/): The rebrand from LeadingAgile to LiminalArc and what it signals. The same company, now positioned around what the market actually needs: AI transformation. - [Introducing LiminalArc (Formerly LeadingAgile)](https://www.liminalarc.co/2025/07/introducing-liminalarc-formerly-leadingagile/): Video introduction to LiminalArc, the rationale for the new name, and the through-line from the founding thesis to the AI transformation frame. ## Optional - [Privacy Policy](https://www.liminalarc.co/privacy-policy/): How LiminalArc collects, uses, and protects information from site visitors. - [LinkedIn](https://www.linkedin.com/company/liminalarcgroup): Company updates, articles, and thought leadership on LinkedIn. - [YouTube](https://www.youtube.com/@liminalarcgroup): Video library, talks, webinars, and short-form explainers. - [RSS Feed](https://www.liminalarc.co/feed/): RSS feed for the LiminalArc blog and resources.