Your New Research Partner
Before you can ask good questions, you need good context. This module shows how AI becomes the BA's most powerful preparation tool — cutting research time dramatically while making you sharper, better-informed, and more credible in every room you walk into.
The BA prep problem — and why it matters
There's a specific frustration most experienced BAs know well. You're assigned to a new engagement — a Guidewire PolicyCenter implementation, a claims transformation, a regulatory compliance project — and you have limited time before your first stakeholder session. You need to get up to speed on the business domain, the current system, the regulatory environment, and the key stakeholders' priorities. You have a day, maybe two.
The traditional approach involves a lot of Googling, reading dense documentation, asking colleagues who've worked in similar contexts, and hoping your prior experience fills the gaps. It works — but it's slow, patchy, and heavily dependent on who you happen to know.
AI changes this entirely. Not by replacing your judgment about what matters — that remains yours — but by dramatically compressing the time it takes to get from "I don't know this context yet" to "I'm informed enough to ask the right questions."
AI doesn't make you an instant expert. It makes you informed enough to have a productive conversation with the actual expert — which is exactly what a great BA does. You're not replacing the stakeholder's domain knowledge with AI-generated summaries. You're arriving at the conversation better prepared to ask the questions that surface what you need to know.
This distinction matters professionally. The BA who arrives at a discovery session and says "I've done some reading on your current Guidewire configuration and I have some specific questions about your data migration approach" is in a completely different position to the one who says "so, tell me about your current system." Same role, same experience level — different preparation quality. AI is what closes that gap, consistently, on every engagement.
AI as a research accelerator — what it actually gives you
Before getting into specific prompts and techniques, it's worth being clear about what AI research assistance actually provides — and what it doesn't. This shapes how you use it and protects you from the failure modes that matter in professional contexts.
Conceptual orientation
Getting a solid working understanding of how Guidewire PolicyCenter works, what FSRA regulates, how a mutual insurance company's policy lifecycle differs from a direct insurer. Fast, reliable, excellent for building a mental model.
Question generation
Given a context, what are the important questions a BA should be asking? AI is exceptionally good at generating comprehensive question sets covering business, technical, regulatory, and stakeholder dimensions — many of which you'd have thought of anyway, but faster.
Gap identification
Given what you know about a project, what are the likely areas of ambiguity, risk, or missing information that a BA should surface? AI can give you a structured gap analysis starting point based on what you describe.
Document synthesis
Paste in a dense requirements document, a previous BA's notes, or a regulatory guidance document and ask AI to summarise the key points, extract the constraints, or identify apparent contradictions. Enormous time-saver.
Domain translation
Translating between business language and technical language is core BA work. AI can help draft explanations of technical concepts for business stakeholders, or translate business requirements into language developers can work with.
What it cannot give you
Current regulatory specifics (verify from FSRA/OSFI/OPC directly), your client's specific systems and processes, undocumented institutional knowledge, and the political context that shapes what stakeholders will and won't say. These require primary sources and direct engagement.
Discovery prep in practice — real prompts, real outputs
The difference between a useful AI research session and a frustrating one comes down almost entirely to how you frame your request. For BA prep work, two prompt patterns are especially productive: the context brief and the question generator. Here's what both look like in an insurance delivery context.
This prompt gives AI everything it needs to produce a genuinely useful briefing document. The output won't be perfect — you'll want to verify any specific version details or feature claims against Guidewire's own documentation — but the conceptual map it produces will be accurate enough to orient you and flag the right areas for deeper investigation.
Don't use the AI-generated questions verbatim. Read them, remove duplicates of things you already know, add questions from your own experience that the AI wouldn't know to ask (because it doesn't know this client's specific history), and reword anything that doesn't sound like you. You're editing a first draft — not reading from a script.
Situation: A BA is assigned to a project reviewing claims handling processes against updated FSRA guidance on auto insurance claims timelines. They need to prepare for a session with the client's compliance team.
Good AI use: Ask AI for a conceptual overview of Ontario auto insurance regulation and how FSRA's supervisory role works. Ask it to generate questions that explore how the current claims process is documented and where the compliance team believes the gaps are. Use this to understand the landscape.
Critical verification step: Do not use AI-provided specific FSRA requirements, timelines, or compliance thresholds in the session. These must come from the actual FSRA guidance documents. AI's training data has a cutoff, regulatory guidance changes, and the consequences of citing incorrect requirements in a compliance context are serious. AI orients you — primary sources inform you.
Understanding unfamiliar systems and domains quickly
One of the realities of consulting in insurance IT is that you regularly encounter systems, platforms, and regulatory frameworks you haven't worked with before. A BA who spent five years on Guidewire may know very little about Majesco or Duck Creek. Someone who's worked exclusively in personal lines may be unfamiliar with commercial lines processes. AI is your fastest path to "informed enough to engage" on any of these.
The AI-accelerated BA prep loop — from unfamiliar context to prepared in under an hour
Platforms and systems worth knowing how to research with AI:
In insurance delivery contexts you'll encounter Guidewire (PolicyCenter, ClaimCenter, BillingCenter), Majesco, Duck Creek, Applied Epic, custom legacy systems (AS/400-based policy systems are still common in Canadian insurers), and various third-party integrations (credit bureau, postal, telematics). For each of these, AI can give you a solid functional and architectural overview that takes hours to find otherwise.
Regulatory frameworks that come up repeatedly in Ontario and Canadian insurance BA work: FSRA (auto insurance, market conduct), OSFI (federally regulated entities), PIPEDA/Bill C-27 (privacy), provincial insurance acts, and increasingly, OSFI's guidance on technology and third-party risk. AI can orient you on all of these — but as noted, specific current requirements must always be verified from primary sources.
The document synthesis use case. One of the highest-value research applications for BAs is pasting existing project documentation into an AI conversation and asking targeted questions about it. A 60-page BRD from a previous project phase, a vendor's technical specification, or a regulatory guidance document — all of these become navigable in minutes rather than hours when you can ask "what are the stated integration requirements in this document?" or "what constraints does this document place on the data migration approach?" or "identify any requirements in this document that appear to conflict with each other."
Module summary — what you can do differently tomorrow
This module is deliberately practical. The goal isn't to understand AI research assistance in the abstract — it's to change how you prepare for your next engagement. Here's what you now have that you can apply immediately.
The context brief pattern
A four-component prompt structure for getting a fast, reliable conceptual overview of any system, platform, or domain — Guidewire, Majesco, FSRA regulation, commercial lines processes, anything.
The question generator pattern
How to use AI to generate comprehensive, themed discovery question sets — and how to edit and personalise the output so it reflects genuine preparation rather than a generic list.
Document synthesis
Pasting existing project documentation into AI to extract summaries, identify gaps, surface TBDs, and generate questions. One of the highest-ROI BA applications of these tools.
The regulatory boundary
AI orients you on regulatory frameworks — it does not replace primary source verification. Specific FSRA, OSFI, PIPEDA, or provincial act requirements must always come from the actual source document.
Module 02 — Requirements at Speed — moves from preparation into the core of BA work: eliciting requirements, writing user stories and acceptance criteria, and using AI to make requirements artefacts faster, more complete, and more consistent. This is where the daily time savings really compound.
You've finished Your New Research Partner. Your progress is saved. Continue to Module 02: Requirements at Speed.