Unified Data
Solutions

Finance leaders no longer need convincing that AI matters. According to a study conducted by leading consulting firm, nearly nine in ten CFOs now say AI adoption in the finance function is extremely or very important, and the share describing their own posture toward AI as conservative has collapsed from 70% in 2020 to just 4% today. The debate has moved on from “should we adopt this” to “why isn't it working the way we expected.” That second question deserves more scrutiny than it usually gets. Across industries, Gartner's research into AI infrastructure and operations found that only 28% of AI use cases fully succeed and meet ROI expectations. Another 20% fail outright. Among leaders who reported a failure, 57% said the root cause was simple: they expected too much, too fast. “AI in finance rarely fails because the technology is broken—it fails because the operating model is.” Furthermore, an independent analysis by the RAND Corporation of enterprise AI initiatives puts the broader enterprise AI failure rate even higher, finding that roughly 80% of projects never deliver their promised business value. Read through a finance lens, that failure pattern can be understood as spanning three broad outcomes: initiatives abandoned before production, initiatives that reach production but underdeliver, and initiatives that run but never recoup their cost. Read these findings together, and a clear pattern emerges: AI in finance rarely fails because the technology is incapable. It fails because organizations run into four structural roadblocks that no amount of additional tooling can resolve on its own. This happens because each roadblock is treated as an isolated problem instead of a symptom of the same underlying gap: the absence of an operating model built for how AI actually needs to be deployed, governed, and budgeted for in a regulated function. ROI fragmentation “According to McKinsey research, large majority of organizations saw no measurable P&L impact from their initiatives” The first roadblock shows up early. Generic, horizontal AI tools are built to be broadly competent, not deeply correct, in finance, ‘broadly competent’ is not the same as reliable. A model that drafts serviceable marketing copy will not, by default, reason correctly through a complex revenue recognition edge case or a non-standard credit memo. The judgment those tasks require is deeply domain-specific, not general-purpose. The data bears this out In awidely cited report issued by MIT Nanda Initiative into enterprise generative AI deployments found that the large majority of organizations saw no measurable P&L impact from their initiatives, while a small minority extracted real, documented value. The differentiator was rarely the underlying model. McKinsey research similarly finds that workflow redesign is one of the strongest predictors of financial value from generative AI. Among 25 organizational attributes examined, redesigning workflows had the largest effect on an organization’s likelihood of realizing EBIT impact from gen AI, reinforcing the idea that value comes less from the tool itself and more from reengineering the underlying process around it. Runaway costs “Gartner's analysis found that organizations are rapidly moving from experimentation to scaled deployment, but are structurally underestimating the financial impact of rising token consumption” The second roadblock is economic, and it is compounding faster than most finance functions have built the governance to track. Gartner's analysis of consumption-based AI pricing found that organizations are “rapidly moving from experimentation to scaled deployment,” but are structurally underestimating the financial impact of rising token consumption. Because developers and business users naturally optimize for speed and convenience over cost, the bills pile up quickly. Left ungoverned, this pattern is materially expensive at scale. Gartner projects that through 2028, at least half of generative AI projects will overrun their budgeted costs specifically due to poor architectural choices and a lack of operational cost discipline, not due to the underlying model pricing itself. This is precisely the kind of risk a traditional software budget was never built to catch. Consumption-based AI spend behaves less like a fixed license fee and more like a utility bill that scales invisibly with usage until the invoice arrives. Without cost governance built in from day one—tracking cost per task, per workflow, per outcome—finance functions frequently discover the economics only after the pilot has already scaled past the point where redesign is cheap. Vendor lock-in “It is highly unlikely that a single AI vendor or model will meet every one of an organization's requirements over time” The third roadblock is strategic rather than financial, though it eventually becomes a cost problem too. Relying on a single frontier model is a structurally fragile position in a market moving this fast: model rankings shift within a single product cycle, providers deprecate versions with limited notice, and outages at a single vendor can take an entire dependent workflow offline. Gartner's own guidance on this point is direct: enterprises should adopt a multi-vendor approach. It is highly unlikely that a single AI vendor or model will meet every one of an organization's requirements over time. Furthermore, hybrid strategies also give finance functions real leverage on token cost as more efficient, open-source, or specialized models enter the market. The lock-in risk compounds specifically in agentic deployments, where every prompt, guardrail, and tool-calling pattern gets hand-tuned to one specific model's behavior. This turn a later migration a substantial engineering effort rather than a configuration change. That is precisely why Gartner projects that by 2028, 70% of organizations building multi-LLM applications will route traffic through a model-agnostic abstraction layer, up from under 5% in 2024. The earlier that abstraction layer is architected in, the cheaper it is to keep future options open. The talent gap “The domain expertise required technical accounting judgment, audit standards, and risk advisory context cannot be substituted with general technical fluency alone.” The fourth roadblock is the hardest to solve with a purchase order, because it isn't really a technology gap, it's a hiring problem for a role that barely exists in the labor market. Deloitte's most recent finance workforce research found that 64% of CFOs and finance leaders have already named at least one technical skill a development priority for their teams, and identified some of the hardest capabilities to find and hire for as cash flow management, financial analysis, budgeting and forecasting, and generative AI experience - in combination, not in isolation. That combination is the actual bottleneck. A strong data scientist without technical accounting depth, or a strong controller without AI fluency, can each get partway to a working solution. Getting the rest of the way requires both perspectives in the same room, wrapped into the same workflow. This is consistent with what Deloitte's broader enterprise research finds across functions: the AI skills gap is now cited as the single biggest barrier to AI integration, ahead of data quality, budget, or governance concerns. In a regulated function like finance, that gap is sharper still, because the domain expertise required technical accounting judgment, audit standards, and risk advisory context cannot be substituted with general technical fluency alone. The common pattern underneath the four roadblocks Treated individually, these four problems look like four separate procurement decisions: pick a more specialized tool, negotiate a better pricing tier, add a backup vendor, or hire a few more tech specialists. Treated as a system, however, they point to the same root cause: Most finance functions are trying to run an AI capability without the operating model that AI capability actually requires. They lack a structured way to match use cases to the right level of specialization, govern consumption cost before they compound, keep architecture portable by design, and pair technical execution with real domain judgment from day one. This is also why maturity, not ambition, is the better lens for thinking about where a finance function actually stands. An organization does not move from “no AI” to “AI-mature” in a single deployment. It moves through a curve-experimentation, first production use case, cost and governance discipline and sustained multi-process advantage. Each of the four roadblocks tends to appear at a different point on that journey: Fragmented ROI is usually an experimentation-stage failure. Runaway cost is a scaling-stage failure. Vendor lock-in is a maturity-stage failure that early-stage teams rarely see coming. The talent gap sits underneath all three, acting as the ultimate governor of how fast an organization can move along the curve at all. The strategic takeaway None of these four walls get solved by adding another point solution to the stack. They get solved by building deliberately, process by process - the operating discipline to match the right level of AI sophistication to each usecase, govern its cost before it compounds, keep the underlying architecture portable, and pair every deployment with the domain expertise the process actually demands. That discipline is not a tool. It is a capability, and it has to be built as one. Pierag's AI & Digital practice works with finance and technology leaders to build exactly this kind of capability: combining technological agility with deep financial subject-matter expertise to guide custom R&D, govern AI consumption cost before it becomes a budget problem, architect for model portability from day one, and close the domain-technical talent gap that generic vendors and generic hires cannot close on their own so that scaling AI in finance is a structured journey, not four separate fires to fight.   Sources Gartner/Salesforce: only 28% of AI use cases in infrastructure and operations fully succeed and meet ROI expectations; 20% fail outright; 57% of leaders who reported failure cited overly ambitious, poorly scoped expectations —https://www.salesforce.com/news/stories/cfos-invest-ai-for-growth/ https://www.gartner.com/en/newsroom/press-releases/2026-04-07-gartner-says-artificial-intelligence-projects-in-infrastructure-and-operations-stall-ahead-of-meaningful-roi-returns RAND Corporation, “Why AI Projects Fail” (2025): approximately 80% of enterprise AI projects fail to deliver promised business value — https://www.rand.org/pubs/research_reports/RRA2680-1.html McKinsey & Company (2025 State of AI research, as cited by Talyx): Among 25 organizational attributes examined, redesigning workflows had the largest effect on an organization’s likelihood of realizing EBIT impact from gen AI  — the-state-of-ai-how-organizations-are-rewiring-to-capture-value_final.pdf Gartner: “Gartner Predicts AI Coding Costs Will Surpass Average Developer's Salary by 2028 as Token Consumption Surges” — https://www.gartner.com/en/newsroom/press-releases/2026-06-24-gartner-predicts-ai-coding-costs-will-surpass-average-developer-salary-by-2028-as-token-consumption-surges Gartner, “10 Best Practices for Optimizing Generative and Agentic AI Costs” (March 2026): through 2028, at least 50% of GenAI projects will overrun budgeted costs due to poor architectural choices and lack of operational discipline (as cited by TrueFoundry) — https://www.truefoundry.com/blog/the-real-cost-of-generative-ai Gartner analyst Max Goss, on avoiding single-vendor AI dependency and adopting a multi-vendor approach — https://www.computerworld.com/article/4188012/too-good-to-be-true-avoid-free-ai-token-offers-or-risk-vendor-lock-in.html Gartner: by 2028, 70% of organizations building multi-LLM applications will use AI gateway/abstraction-layer capabilities, up from under 5% in 2024 (as cited by Swfte AI) — https://www.swfte.com/blog/avoid-ai-vendor-lock-in-enterprise-guide Deloitte, “Finance Workforce Strategy in the AI Era” — https://www.deloitte.com/us/en/what-we-do/capabilities/finance-transformation/articles/finance-workforce-strategy-ai-era.html Deloitte, “CFO Guide to Human Capital Trends: AI & Human Collaboration” — 87% of CFOs predict AI adoption in finance will be extremely or very important; 64% named at least one technical skill a development priority — https://www.deloitte.com/us/en/what-we-do/capabilities/finance-transformation/articles/cfo-guide-human-capital-trends-ai-human-collaboration.html

In Brief Following a large acquisition, the client inherited 2,000+ unstructured Vendor Service Level Agreements across a highly regulated industry with no visibility into what they contained or what risk they carried Pierag deployed SmartXtract's contract analysis agent: a customised, AI-enabled workflow combining automated data extraction, human expert validation and live data visualisation Planned review time reduced by over 60%. The client gained full portfolio visibility and new strategic capabilities across vendor negotiation, benchmarking and risk management. The Situation When a large enterprise acquires a business entity, it also acquires everything that business has signed. In this case, that meant over 2,000 Vendor Service Level Agreements — unstructured in format, scattered across repositories, and never comprehensively reviewed. Operating in a highly regulated industry made this more than an operational headache. Outdated or expired contracts, non-compliant contractual terms, buried penalty clauses, rebate entitlements, and auto-renewal provisions created significant legal, regulatory, and financial exposure. With no central repository or common taxonomy, and contracts written in highly specialised industry language, the legal and procurement teams had no practical starting point. Manual review was not a viable option. The volume was too large, the timeline too short, and the cost of a missed clause too high. The Challenge The Approach Pierag deployed SmartXtract's contract analysis agent configured for the client's industry and contract portfolio through a six-step workflow designed to move every agreement from unstructured document to structured, actionable intelligence. 01 · Secure Contract Upload All agreements ingested through a secure layer with full data confidentiality and audit traceability from day one. 02 · AI-Enabled Data Extraction Proprietary models extracted and classified key fields across all agreements simultaneously - parties, term dates, Service Level Agreement obligations, penalty clauses, rebate terms, renewal conditions and niche industry-specific terminology. 03 · Human-in-the-Loop Review Complex and ambiguous clauses routed to subject-matter experts ensuring accuracy where contractual nuance cannot be left to pattern recognition alone. 04 · Automated Dual-Model Validation Every output cross-checked for consistency and completeness. Exceptions flagged before results moved forward. 05 · Data Protection The tool was built on a stateless architecture ensuring no data left the client’s system during processing. 06 · Standardised Output All contract data delivered in a uniform, structured format ready for immediate integration into legal, procurement and finance workflows. 07 · Data Visualisation Extracted insights surfaced through a live dashboard giving leadership visibility into the full contract portfolio.   The Impact 60% reduction in planned review time from a projected six months to under two. Beyond speed, the engagement delivered capabilities the client did not have before: Operational  Faster contract reviews at scale, improved transparency across the acquired portfolio, reliable extraction of critical clauses and terms, standardised outputs for analysis and reporting, and significantly reduced dependency on manual review — freeing the team for strategic and risk-focused work. Strategic Vendor negotiation leverage: At renewal, the procurement team can access the dashboard, view the exact penalty framework against vendor performance and negotiate from evidence rather than assumption. Service Level Agreement benchmarking : Similar vendors - catering, IT services, logistics can now be compared on contractual terms, identifying who holds the most favourable conditions and where renegotiation is warranted. Concentration risk visibility : Leadership can see whether too many critical operational dependencies are tied to a single vendor group — and whether the contractual terms governing those relationships are adequate. Standardisation gaps identified : Older contracts lacking modern indemnification or penalty clauses have been surfaced and queued for remediation — giving legal a clear, sequenced workload. From reactive to proactive: Instead of digging through documents after a vendor failure, the team can see risk profiles instantly - and act before issues escalate. About SmartXtract SmartXtract is Pierag's AI-agentic platform hosting specialised finance agents built for real use cases — contract analysis, technical accounting, reconciliations, reporting and more. Each agent mirrors the workflows of finance professionals by extracting data, interpreting standards and delivering structured outputs. Developed on Azure Platform, securely hosted and easily scalable, SmartXtract integrates with your data to provide accurate, actionable insights in real time.   Discover how Pierag's AI & Digital lab helps organisations convert unstructured documents into actionable intelligence.

Artificial intelligence is no longer a fringe innovation topic or a limited pilot initiative. It has moved firmly into the enterprise mainstream, and the market conversation has shifted accordingly: from experimentation to scale, governance, operating models, and measurable value creation. That shift matters because adoption is no longer the real test of maturity. Most enterprises today can access leading models, license copilots, launch pilots, and introduce AI-enabled tools into selected functions. Yet the presence of AI in the technology stack does not, by itself, improve business performance. The more important question is whether AI has been embedded in a way that meaningfully improves how work gets done. This is where many enterprise AI programs begin to lose clarity. Attention often centres on model selection, tool comparisons, or the promise of the latest platform release. Those decisions are relevant, but they are not usually what determines long-term value. In practice, the more difficult and more consequential challenge is execution: defining where AI belongs within a business process, where conventional automation is more effective, where human judgment must remain central, and how all of it is governed at scale. Recent enterprise experience has made this distinction increasingly visible. Many of the most instructive AI stories are not about whether the technology works in principle. Instead, they are about cost overruns, unclear returns, weak process fit, inconsistent usage, and the difficulty of scaling tools that were introduced without sufficient operational discipline. That is why enterprise AI strategy should not begin with access to technology. It must begin with the design of work. Shifting the Focus: From Access to Execution A more effective starting point is the business workflow itself. To build a grounded, impactful AI roadmap, leaders must step back from the technology and begin by asking critical diagnostic questions: Process Centrality: Which processes are genuinely central to our operational performance? Friction Points: Where do teams currently lose time to review, rework, handoffs, or fragmented information? Cognitive Demands: Which activities depend most heavily on pattern recognition, contextual interpretation, or complex exception handling? Value Leverage: Where would better support improve quality, consistency, customer experience, or the speed of decision-making? These questions tend to produce a much more grounded roadmap than a technology-first approach. They also lead to an essential realization: not every business problem requires AI, and not every step within an AI-enabled process should be handled by AI. Deconstructing the Workflow Enterprise workflows contain distinct categories of work, and each category demands a different tool. Some steps are deterministic and governed by stable rules. Others are repetitive and process-driven. Some involve ambiguity, unstructured information, or complex contextual interpretation. Others require legal accountability, commercial judgment, or strict compliance oversight. Treating all of these as the same kind of problem is one of the most common errors in enterprise AI design. Strong AI programs are rarely built by maximizing the amount of AI in a process; they are built by assigning the right capability to the right type of task. The vendor invoice process offers a highly practical illustration of this multi-layered framework. Consider a multinational enterprise managing thousands of global suppliers. Seeking a quick win, leadership deploys a generic, off-the-shelf AI co-pilot to automatically read and approve all incoming invoices. In reality, the initiative quickly derails. The generic AI hallucinates on standard tax fields because it does not understand the firm's strict internal data boundaries. It wastes expensive computational power simply routing a PDF from a manager to a VP. Worst of all, it mistakenly approves a disputed, high-value transaction because it lacks the commercial context of an ongoing vendor lawsuit. This failure occurs because the enterprise treats the entire workflow as an "AI problem." In reality, to succeed, the process must be deconstructed into a layered operating model: Rules with Predictable Outcomes: Validating invoice data against purchase orders and tax requirements is entirely rule-based. A classic rules engine is the most cost-effective and reliable tool here. Automating Repetitive Tasks: Procedural steps, including routing approvals and updating ERP systems, are highly repeatable. Standard workflow automation is best suited to these status-based actions. AI Where Context and Judgment Are Required: AI becomes highly valuable during exceptions—unusual charges, incomplete documentation, or subtle inconsistencies. Here, AI can analyze unstructured supporting material, highlight anomalies, and assist a reviewer in narrowing down the issues. Humans When Accountability Counts: Human oversight remains strictly necessary where commercial judgment, regulatory sensitivity, supplier disputes, or high-value decisions require a level of accountability that cannot be delegated to an algorithm. This layered operating model is far more effective than the blanket idea of “AI everywhere.” It respects the unique strengths of rules, automation, AI, and human expertise. Custom Architecture vs. Generic SaaS Because strategic enterprise workflows require this precise, multi-layered coordination, managing the handoffs between strict business rules, standard automation, and cognitive AI assistance requires a cohesive, tailored orchestrator. Generic, off-the-shelf software rarely has the inherent flexibility to stitch these four layers together seamlessly. This raises a critical question for enterprise leaders: when is a standard platform sufficient, and when is custom development justified? The most effective standard operating procedure (SOP) relies on a simple distinction: Core versus Context. Context Workflows (Buy/Standard SaaS): These are non-differentiating processes that every company handles similarly—such as standard payroll processing, routine expense categorization, or basic accounts payable routing. For these, standard, off-the-shelf platform tools are completely sufficient. There is no strategic value in reinventing the wheel. Core Workflows (Build/Custom Orchestration): These are the proprietary processes where an enterprise actually wins its market—whether that is an investment bank's proprietary M&A valuation model, an consulting firm’s technical accounting expertise, or a multinational's complex forecasting engine. This challenge has become top-of-mind as adoption timelines compress. Research from the Wharton School and GBK Collective indicates that generative AI is fast-tracking into the core of the enterprise, with decision-makers increasingly shifting budgets from experimentation to integration. Yet, as adoption accelerates, the gap between high-performing and average organizations becomes clearer. McKinsey’s research indicates that while AI use is now widespread, the ability to translate that use into actual business impact remains highly uneven. Crucially, high-performing organizations are far more likely to redesign workflows fundamentally to support this multi-layered reality, rather than simply overlaying AI onto existing, broken activity. Enterprise value is created not when AI is added on top of work, but when work itself is redesigned to use AI appropriately. In specialized, "Core" environments, durable value typically comes not from buying another off-the-shelf license, but from configuring bespoke solutions that align perfectly with the unique operating model of the business. Governance and the Power of Human Augmentation To support this bespoke architecture, an enterprise operating model must prioritize governance and human enablement from day one. Cost control, usage discipline, data boundaries, and security guardrails cannot be added as an afterthought. The growing emphasis in enterprise research on production readiness and ROI measurement reflects exactly this concern. For instance, ISG’s reporting focuses heavily on spending trends, governance, and scaling challenges, while other market research increasingly evaluates AI not by its novelty, but by its deep integration and structural guardrails. Crucially, those guardrails are not just technical—they are human. One of the most persistent misconceptions is that AI's primary value lies in replacing human labor. In reality, the International Monetary Fund’s (IMF) analysis of labor exposure continually emphasizes complementarity—the immense potential of technology to work alongside people, amplifying their capability rather than substituting it. AI does not create enterprise value simply because it is available. It creates value when employees are trained and empowered to co-pilot with it: Understanding precisely where the technology adds cognitive leverage. Knowing exactly where its outputs must be challenged or verified. Recognizing where over-reliance would introduce unnecessary operational risk. This is particularly relevant in high-stakes functions like finance, legal, procurement, and risk, where work constantly balances structured process and contextual judgment. In these environments, staff education is not a secondary HR workstream; it is a core part of the operational control framework and the ultimate engine of productivity. Strategic Execution: The Path to Lasting Value The broader business case for this disciplined approach is becoming impossible to ignore. Recent market research highlights a widening performance gap between organizations that merely acquire tools and those that build the operational infrastructure to support them. Oxford Economics’ work on enterprise AI maturity, for example, demonstrates that sustainable value is tied directly to deep operational integration rather than simple access. This is reinforced by McKinsey’s findings, which establish a direct link between fundamental workflow redesign and actual value capture. Ultimately, this execution gap translates into a financial one: as IMD’s maturity research points out, a significant performance and margin divide is opening up between operationally mature enterprises and those still struggling to scale their pilots. The implication for enterprise leaders is straightforward. The long-term winners in the AI era are unlikely to be the organizations that deployed the greatest number of tools or announced the largest number of pilots. They are more likely to be the ones that: Identified the right strategic workflows. Intelligently combined rules, automation, AI, and human oversight. Built robust governance frameworks from the outset. Trained their workforce to interact with these capabilities with disciplined, empowered skepticism. AI adoption may open the door, but disciplined execution determines whether that investment translates into durable business value. Sources International Monetary Fund (IMF)- sdnea2024001.pdf ISG (Information Services Group)- isg-one.com/docs/default-source/default-document-library/2025-isg-state-of-enterprise-ai-adoption-r… Wharton School / GBK Collective- ai.wharton.upenn.edu/wp-content/uploads/2025/10/2025-Wharton-GBK-AI-Adoption-Report_Full-Report.pdf Impact AI Series | Oxford Economics IMD Business School- Companies leading in AI adoption use it as a catalyst for reinvention - IMD business school for man…