A Fortune 500 consumer product testing company authors thousands of test protocols that define how products are evaluated for safety, performance and consumer experience. Prognos Labs designed and delivered a GenAI-powered protocol authoring platform that moves this work out of spreadsheets and documents and into a governed, AI-assisted software workflow, achieving 85% ROI within three months.

The Challenge

Protocol authoring relied on Excel test sheets and Word documents, with formats that changed by product category. Knowledge sat in individual files and in the heads of experienced authors, which made the work slow and hard to standardize. The company needed to:

• Reduce the time authors spend drafting and revising each protocol.

• Reuse prior protocols and product knowledge instead of starting from scratch.

• Bring real-world consumer feedback, such as product reviews, into protocol design.

• Track changes and versions reliably across a large product hierarchy.

• Meet enterprise security and network standards from day one.

The Solution

Prognos Labs built a platform around four capabilities, all running on the client's Azure and Microsoft environment.

AI-Assisted Authoring

• A chat interface where authors describe what they need and receive grounded protocol drafts.

• An extraction-first RAG architecture that retrieves from existing protocols and product data, so outputs reflect the company's own standards.

Consumer Insight Pipeline

• A scheduled pipeline, run on Azure Functions, that collects user reviews from external sources.

• Extracted insights feed directly into the authoring workflow, tying test design to real consumer concerns.

Version Control & Product Hierarchy

• Every protocol change is versioned, so teams can see what changed, when and why.

• Protocols are organized by product hierarchy, replacing category-specific spreadsheet formats with one consistent structure.

Enterprise-Grade Security

• Least-privilege PostgreSQL roles, with separate deployer and runtime identities backed by Entra.

• Secrets managed in Key Vault, monitoring through Application Insights, and a completed network and security review.

Approach & Timeline

The engagement ran for three months, moving from process discovery to a platform in daily use.

Month 1 - Discovery & Design: mapped existing authoring workflows, defined the product hierarchy and requirements, and designed the RAG and Azure architecture.

Month 2 - Core Build: delivered the chat interface, version control, product hierarchy and the review-collection pipeline, along with the security configuration.

Month 3 - Rollout & Optimization: released to authors, tuned retrieval and output quality from real usage, and measured results.

The Results

• 30 minutes saved per protocol, as drafting starts from grounded, reusable content.

• 25% productivity boost for the authoring team.

• 85% ROI within three months of the engagement.

• One consistent, versioned system of record in place of scattered spreadsheets and documents.

Why It Worked

The platform kept authors in control. AI handles retrieval and first drafts, while people review, edit and approve. Grounding every output in the company's own protocols and product data kept the content trustworthy, and building security and version control in from the start meant the platform passed enterprise review without rework.

About Prognos Labs

Prognos Labs is an AI consulting and product development firm that partners with founders and enterprise teams to design, build, and ship AI-powered products, from early-stage concepts to production-grade platforms.