Nitish Shrivastava, CTO, Engineering Hyper Productivity, Persistent Systems
The Frontier of Institutional Capability
The frontier of AI has largely been defined outside the enterprise, through the model: deeper reasoning, longer context, additional modalities and greater autonomy. Inside the enterprise, capability enters a different equation.
Here, intelligence encounters systems formed over decades, where software carries the memory of the business. Business rules live across code, data, exceptions, operating procedures, release histories and decisions that may never have been written down. Architecture is inherited as often as it is designed. Regulatory obligations, data boundaries, finite budgets and live operations shape every act of change.
What Institutions Carry
Across industries, from banking and telecommunications to public safety, defense and mature software platforms, institutions carry consequences that go far beyond technical performance: fiduciary and regulatory accountability, national continuity, human safety, sovereign responsibility, and decades of accumulated decisions and exceptions.
A model may perform brilliantly in isolation and still be unsuitable for the work.
The relevant measure is larger than model capability: it is the weight of consequence an intelligent system can carry while institutional intent, security and economic control remain intact.
The SASVA Position
"Enterprise intelligence is measured not by what a model can do in isolation, but by what an institution can safely and repeatedly accomplish with it. SASVA brings institutional intent, technical reality, governance and consequence into one executable system, enabling AI to operate where reliability, sovereignty and control cannot be compromised. SASVA is built for that measure, from the institution outward, placing institutional priorities, constraints and consequences ahead of the model."
That is the measure SASVA is built to meet.
For Agentic SDLC, the AI Coding Harness Matters More Than the Coding Model
- C.A. Swan
- 24 March 2026
Software engineering leaders should prioritize engineering a harness layer before scaling agent adoption by selecting a harness product (such as Claude Code, OpenAI Codex or Kiro) and assigning platform engineering teams to own context management, real-time backpressure and verification gates.
Strategic Planning Assumption
Issue
AI coding tools have achieved near-universal adoption across software engineering organizations, 1 and teams increasingly apply AI beyond code generation to testing, documentation, deployment and other SDLC activities. 2
But increased development velocity has not translated into equivalent improvements in software quality. Only 53% of leaders rate their teams effective at meeting software quality standards, 1 and only 39% of engineering leaders and teams report reduction in technical debt of newly generated code as a result of using AI tools in SDLC. 2 More code is being produced and shipped faster, without the governance controls required to maintain quality at scale.
The AI coding harness addresses this gap. The harness operates at two levels.
- The harness product is an agentic coding platform (such as Claude Code, OpenAI Codex, OpenCode, Cursor, Google Antigravity or Kiro) that provides primitives for context delivery, enforcement, verification and model routing.
- The harness layer is the organizational configuration built on top of the product: rules files, coding standards, enforcement hooks, skills, verification gates and managed settings that a platform team engineers and maintains.
Every agentic coding platform provides baseline harness capabilities, but organizational configuration determines whether the harness enforces standards specific to a given codebase. Together, these components govern how AI agents generate and validate code before it reaches the codebase.
Software engineering leaders should implement harness infrastructure to ensure the productivity gains from AI coding tools do not translate into higher defect rates and technical debt. Read more
Persistent Systems
Our understanding of the March 2026 Gartner research, For Agentic SDLC, the AI Coding Harness Matters More Than the Coding Model, is that context engineering, enforcement hooks, and verification gates play a critical role in enabling software quality in agentic development environments. We believe the principles described in the research align with SASVA's approach to enterprise execution architecture, where the coding harness is one component of a broader lifecycle framework.
Gartner states: "By 2028, organizations with AI coding harness infrastructure will experience 60% fewer postdeployment defects than those without." Based on our reading of the research, this reinforces the importance of establishing governance, context, verification, and operational controls around AI-assisted software delivery.
| OUR GARTNER RESEARCH THESIS | SASVA’s ROLE |
|---|---|
| Organizational context should shape the model before generation. | Design authority should move upstream into an architect-approved, executable representation of the initiative. |
| Enforcement should interrupt nonconforming work during generation. | The system boundary should span assessment, product intent, planning, analysis, architecture, engineering, testing, migration, release, service, and support. |
| Verification should require executable evidence before completion. | Context should become a governed, versioned estate with provenance, lineage, staleness propagation, consumption controls, and phase-level enforcement. |
| Institutional knowledge encoded in the harness compounds across model changes. | Security, model selection, execution environments, and token budgets should be managed as part of a unified design framework, with memory and evidence preserved across long-running, multi-agent, multi-environment work. |
To us, the line of agreement is clear: the system around the model determines production quality. The line of extension is equally important: enterprise work begins before coding and continues after deployment. SASVA turns this harness principle into a full-lifecycle engineering architecture.
Source & Attribution
Disclaimer
GARTNER is a trademark of Gartner, Inc. and/or its affiliates. Gartner does not endorse any company, vendor, product, or service depicted in its publications and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner publications consist of the opinions of Gartner's business and technology insights organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this publication, including any warranties of merchantability or fitness for a particular purpose.
Learn more about SASVA here.

