Case studies / AskGXP

AI PRODUCT ENGINEERING · LIFE SCIENCES

AskGXP - an AI assistant for GxP documentation

An AI product engagement for a pharmaceutical documentation platform. It combines retrieval-based AI, real-time streaming and document export in one web application.

AskGXP

Pharma · AI GxP documentation assistant


5

Focused AI Agent in one workspace

NetSet Software AI product engineering partner

At a glance

Domain

Pharmaceutical GMP / GLP

Focus

AI document generation

Engagement

Fixed-cost product engineering

Timeline

8 months, from discovery to launch

The partnership

A specialist domain. A practical AI product.

AskGXP came to NetSet Software with a clear aim: help pharmaceutical professionals produce GxP documentation faster, without losing the depth that regulated work needs.

Documents such as SOPs, User Requirement Specifications and validation protocols take a long time to write and depend on deep knowledge of GMP and GLP practice. The client wanted an assistant that drafts these documents from a trusted domain knowledge base, rather than a general-purpose chatbot.

NetSet Software designed and built the product end to end: the AI service, the web application, user accounts and the subscription experience.

The product

A guided assistant for GxP work.

AskGXP brings a set of focused assistants into one workspace. Users choose a department and topic to generate a structured Standard Operating Procedure. They can draft User Requirement Specifications and validation protocols, and use Guided Learning and an AI chat for GMP and GLP questions. Answers stream in as they are written, and finished documents can be exported to Word for review and editing within the organisation’s own quality process.

Engineering approach

Engineering around the product experience.

01 — Ground the answers in domain knowledge

Responses draw on a curated library of pharmaceutical guidance, so drafts reflect GxP practice and not just general AI knowledge.

02 — Make long documents feel fast

SOPs and protocols are long. Real-time streaming lets users watch the document form section by section instead of waiting on a blank screen.

03 — Fit the way quality teams work

Generated content exports straight to Word, so teams can review, edit and route drafts through their existing approval process.

04 — Build a product, not a prototype

User accounts, history, subscription plans and an admin area for managing templates make AskGXP a complete SaaS product from day one.

The AI layer

Retrieval-augmented generation. Streamed in real time.

Pharmaceutical source documents are processed, split into searchable passages and indexed in a vector store. When a user asks a question or requests a document, the platform pulls out the most relevant passages and passes them to a large language model with prompts designed for GxP documentation. The response streams back to the browser over WebSockets.

“Domain-grounded AI drafts, delivered in real time, ready for expert review.”

Technology

Python FastAPI LangChain OpenAI FAISS Hugging Face embeddings Next.js React Chakra UI Tailwind CSS Firebase Stripe

The engagement model

Product engineering. Fixed cost. Four months.

Scope agreed up front at a fixed price, so the client had a clear budget before development began

Four months of two-week sprints, with a working build reviewed by the client after each sprint

Outcome

A working AI platform for pharmaceutical documentation.

AskGXP launched as a subscription web application that lets pharmaceutical professionals move from a blank page to a structured first draft in minutes. Documents that take several hours to draft by hand become a reviewable first draft in under five minutes

What this gives a product business

What this gives a product business.

Domain-specific AI: AI built around the client’s own knowledge base and use cases, not a generic chat interface.

Faster time to value: Users get structured, usable drafts for complex documents from their first session.

Revenue-ready from launch: Subscription plans and payments are built into the product, ready to monetise from launch.

Room to grow: A modular assistant structure, so new document types can be added without rebuilding the platform.

A practical route from domain expertise to AI software.

AskGXP shows how specialist knowledge can become a focused AI product that works within professional standards: grounded in trusted sources, quick to use, and built to support expert review rather than replace it.

AI RAG LLM Pharma GxP SaaS Next.js FastAPI

Turn your domain expertise into an AI product.

We help businesses build AI applications grounded in their own knowledge and workflows.

Discuss your AI product

Let's discuss your engineering plan

Share a few details. We'll get back to you shortly.

What would you like to discuss? *

Your details stay confidential · NDA available before detailed discussion