Accenture and WaveMaker have announced a strategic partnership to help mid-market organizations modernize applications using WaveMaker’s agentic AI platform, targeting companies with annual revenue up to $3 billion that lack the in-house expertise or budget for large-scale digital transformation.
Instead of talking about signed agreements in its announcement issued on Tuesday, WaveMaker states that this “strategic intent” focuses on the market problem/proposition (reliable, cost-effective AI-assisted outcomes for a certain size of customers, solely for app modernization and new applications, and a platform-led services play).
For instance, this week’s news about Uber is emblematic of the current state of AI coding, WaveMaker said. Yet this announcement opens the door to a range of joint GTM and client-delivery motions using WaveMaker’s agentic platform, the company said.
Accenture is reselling, integrating, and co-developing solutions with WaveMaker. WaveMaker’s parent company (Pramati) has had a relationship with Accenture for over five years, since Accenture acquired Imaginea (another portfolio company of Pramati).
Two passes, one guarantee: No hallucinations shipped
The collaboration combines Accenture’s software engineering services with WaveMaker’s code-generation platform, which the company describes as an “industry-first 2-pass architecture” — a deterministic two-stage generation engine designed to produce secure, scalable web and mobile applications from Figma design files and natural-language prompts.
2-pass overcomes the inherent stochastic nature of AI code generation and delivers deterministic AI outcomes at AI speed without breaking the budget.
WaveMaker CEO and Co-founder Vijay Pullur tells The New Stack, “2-pass overcomes the inherent stochastic nature of AI code generation and delivers deterministic AI outcomes at AI speed without breaking the budget. This matters to any enterprise customer, and more so in regulated industries.”
WaveMaker targets regulated industries, where hallucinated or non-compiling code isn’t just an inconvenience. The 2-pass approach is meant to address that: the platform generates code, then verifies it against architectural guardrails before output, rather than shipping whatever the model produces on the first pass.
“Final code is generated in the second pass using a deterministic code generator that is not prone to AI inaccuracies,” Pullur said. “Any hallucination or inaccuracies, if any, can occur only in the intermedia WaveMaker Markup Language (meta model) generation that can be easily rectified by human intervention using our visual canvas.”
The first pass generates a meta-model, and the final code is generated in the second pass using a deterministic code generator that is not prone to AI inaccuracies.
The market nobody owns
While the solution can apply to any market, the problem is acute in companies of the sub-$3B size that are compelled to move super-fast with modernization and transformation, but don’t have the necessary budgets, expertise or time. For WaveMaker, this is also an underserved market with specific attributes, Pullur tells The New Stack.
The partnership is centered on a specific gap Accenture has identified in the market. Large enterprises have transformation budgets and dedicated engineering organizations. Small businesses often have simpler needs. Mid-market companies — complex enough to need real software, constrained enough to feel the cost — tend to fall between the two. Aging systems, SaaS sprawl, and rising competitive pressure are the pressure points Accenture and WaveMaker say they’re targeting.
“Organizations need to modernize and innovate quickly to keep pace, yet many still face constraints around cost, complexity, and in-house expertise,” said Kishore P. Durg, senior managing director in Reinvention Services at Accenture, in a statement. “This collaboration enables our clients to accelerate application modernization and deliver new digital capabilities while maintaining control over costs and reducing operational risk.”
WaveMaker’s Pullur positioned the deal as validation of the company’s architectural approach. “Being in the business of creating a developer-focused agentic AI platform means we have always believed in delivering sound architecture and modern, scalable digital experiences for our enterprise customers,” he said in a statement.
A crowded field, a specific bet
The announcement comes as the agentic AI application development market is getting crowded. OutSystems, Salesforce, and ServiceNow are all pushing into AI-assisted enterprise app generation, and Microsoft and Google are embedding similar capabilities directly into their development toolchains. WaveMaker’s differentiator claim is the 2-pass deterministic engine and its focus on long-lived, team-scale applications — as opposed to the throwaway prototypes that vibe coding tools tend to produce.
WaveMaker sits in a unique space of a hybrid agentic AI system (not a tool) for modern web and native mobile applications that’s enterprise-ready out-of-the-box — combining LLM (model-agnostic) generation with assembly for design-led classy, customer-facing applications; with 2-pass code generation and 12-factor app architecture built-in and bimodal developer experience (chat, visual), Pullur said.
The system is built for early-career teams in midmarket businesses who need a ready-made AI system that delivers reliable results from day one, meeting their teams’ skills, patterns, and tools where they are. The target tech stack is React/Angular for web and React Native for mobile with Java/Spring/Hibernate in the backend, he said.
WaveMaker customers include Fidelity National Information Services, AT&T, Nokia, and Blue Yonder, as well as companies in the energy, retail, healthcare, and insurance sectors.
Guardrails the platform enforces
Meanwhile, regarding guardrails, “When you roll out a developer tool like Claude Code or others, it is very difficult to enforce application architecture, security, design, and user experience compliance across all developers and teams in the company using the tool,” Pullur said.
However, “WaveMaker makes these guardrails possible. Some of them are system-enabled, for example, all applications developed should comply with accessibility features, which is system-enforced,” he explained. “Similar is security-related compliance. Things like sticking to design systems and design tokens across the company for all projects/developers can be configured and enforced in WaveMaker Studio. In summary, the guardrails are design tokens, pre-built components, WaveMaker Markup Language, The Twelve-Factor App architecture baked in, OWASP compliance, and more.”
The post Why Accenture and WaveMaker are betting on agentic AI to close a $3 billion software gap appeared first on The New Stack.