![\"Accelerate](\"https:\/\/www.jamasoftware.com\/media\/2026\/05\/Accelerate-Compliance-and-Requirements-Review-with-AI-2.png\")
<\/h2>\n<\/h2>\nReviewing requirements and their associated compliance material is an essential component of developing a quality product. However, review time can often be consumed by arguing over the minutiae of syntax, grammar, and other secondary aspects of a requirement, rather than the technical subject matter. For many teams, this means their subject matter experts (SMEs) are trapped in debate rather than informing design and progressing the development cycle. With the advent of Large Language Models (LLMs)<\/a> and Model Context Protocol (MCP)<\/a>, the engineering industry is on the verge of optimizing away from these debates and focusing on critical technical progress instead. By tactically prompting and developing tools that can provide a first-pass review, engineering teams can:<\/p>\n Artificial Intelligence (AI)<\/a>, particularly Generative AI, has modernized rapidly. Engineering teams are now deploying personas and agents into their workflows. The use of AI provides many clear benefits; however, there are also looming pitfalls. I have personally noted, while working with customers and in my engineering career, that using a deliberate method that forces the human back into the loop provides more effective results than directly allowing AI to modify requirements.<\/p>\n This is where Jama Connect\u2019s integration with an LLM through MCP improves the engineering workflow. Engineers can develop standardized prompts and personas to review data. With the data reviewed and AI recommendations provided, junior engineers can step back in and make improvements prior to discussing the technical details with SMEs. Therefore, improving the requirement\u2019s first draft quality. Simultaneously, this approach reduces the cost of creating quality requirements by decreasing the reliance on SMEs early on in the process.<\/p>\n Throughout engineering, there are common roles and expertise that apply to requirements development. For example, there are known syntactical rules posted by many organizations for how to author a requirement. Rules from each of these sources can be applied systematically to help review requirements. Thus, it reduces debate over whether the requirement is syntactically correct.<\/p>\n Similarly, many roles throughout engineering are well defined by guiding documents, standards, and best practices. Most of these documents are public and accessible via the internet. Therefore, it is straightforward to develop personas that mimic an SME from their respective domain of expertise. These personas can be as simple as textual prompt files or as complex as logic-based scripts that are run and utilized by an AI integration into Jama Connect. For instance, a mechanical engineering persona is informed by NASA-STD-5001 and MIL-STD-1522<\/a>, among a handful of others, to produce meaningful recommendations.<\/p>\n When authoring a persona, teams can do this natively in Jama Connect as a specific Item Type that includes the response format, list of documents the persona is built around, and other pertinent details. By integrating this information directly into the project where the requirements reside, teams are able to assess and version control each persona and related information. With personas developed, teams can rapidly deploy and redeploy canned expert reviewers to provide recommendations to their requirements. My recommendation is to pair these personas with a custom field within the requirement items. This allows Jama Connect users to see the recommendations made per version of the requirement.<\/p>\n As depicted above, the recommendations from the AI review are simple to understand and actionable. An engineer can process an update to the requirement and execute an additional run of the deployed personas to confirm the updates are satisfactory. Users can feel confident in early drafts of requirements due to this feedback and confirmation cycle. This also enables quick iterations prior to requiring time from SMEs.<\/p>\n A key objective of this approach is to reduce AI hallucinations by directly modifying technical engineering requirements that guide complex design, integration, and test. By doing so, teams reap large benefits through the recommendation cycle, while also minimizing the chance of a late failure due to misguided AI inputs.<\/p>\n The goal of this approach is to reduce time spent on initial requirements drafting. Doing so allows teams to focus their resources on early prototyping, compliance assessments, and other typical early-stage engineering efforts that often become truncated due to requirements development and analysis. As functionality like this progresses, new ways to ensure accuracy and reduce overhead will continue to arise. Keep an eye out for continued explanations and approaches to improve your requirements development and review with Jama Connect and AI.<\/p>\n\n
Hone Engineering Expertise with Personas with AI<\/h3>\n
![\"Prompt](\"https:\/\/www.jamasoftware.com\/media\/2026\/05\/PROMPT-BASED-JAMA-CONNECT-ANALYSIS-SCHEMATIC-5.png\")
<\/p>\n
\nRELATED: <\/strong>Accelerate AI-Driven Development with Jama Connect MCP\u2122<\/a><\/span><\/span><\/h4>\n
\nRapidly Deploy AI into Requirements Development Workflows<\/h3>\n
![\"Jama](\"https:\/\/www.jamasoftware.com\/media\/2026\/05\/JC-Doc-View-985x1030.png\")
<\/p>\n
\nRELATED: <\/strong>Buyer\u2019s Guide: How to Select the Right Requirements Management and Traceability Solution<\/a><\/span><\/span><\/h4>\n
\nFinal Thoughts<\/h3>\n