Business Analyst

Mission

Turn business intent into requirements engineers can build and the Quality Engineer can verify. The BA is the bridge between the why (stakeholders acting as Product Owner) and the how (engineering), producing user stories and acceptance criteria that leave no room for guesswork. At EmeSoft the BA is the primary day-to-day interface to the stakeholders.

Responsibilities

Gather requirements from stakeholders, users, and existing systems.
Write user stories and acceptance criteria (Given/When/Then).
Model processes and data: flows, state diagrams, data dictionaries.
Maintain traceability from business need → story → test.
Lead backlog refinement with the stakeholders (PO) and Architect.
Clarify requirements for developers and testers throughout the sprint.

Inputs

Input	From
Epics, priorities, value hypotheses	Stakeholders (Product Owner)
Domain knowledge & constraints	Stakeholders, SMEs
Technical feasibility input	Architect/Dev Lead
Existing system behavior	Codebase, docs, analytics

Outputs

Output	To
User stories + acceptance criteria	Dev Team
Process / data models	Architect, Developers
Requirement traceability matrix	Quality Engineer, Stakeholders (PO)
Clarifications & decisions log	Whole team

AI tools & how we use them

The BA is one of the heaviest and highest-leverage AI users in the process.

Claude / ChatGPT (story & AC drafting): convert raw notes, transcripts, or a feature idea into structured stories with Given/When/Then criteria. Prompt: “Turn this requirement into 3 user stories with testable acceptance criteria in Given/When/Then; list edge cases I may have missed.”
AI for edge-case discovery: explicitly ask the model to enumerate boundary, error, and negative scenarios - AI is excellent at surfacing cases humans skip.
Mermaid / diagrams.net AI: generate process and state diagrams from a textual description, then refine. (These docs use Mermaid for exactly this reason.)
Requirement quality checks: ask AI to review a story for ambiguity, missing criteria, or untestable language before it enters a sprint.
Confluence / Notion AI: summarize long requirement docs and keep specs consistent.

flowchart LR
    N[Raw notes / transcript] --> AI[AI drafts stories + ACs + edge cases]
    AI --> BA[BA reviews, corrects, adds domain nuance]
    BA --> R{Meets Definition of Ready?}
    R -->|no| BA
    R -->|yes| BL[(Refined backlog)]

Collaboration

The BA partners with the stakeholders (PO) (refining the what) and the Architect (checking the how is feasible), and serves Developers and the Quality Engineer with clarifications. The BA usually facilitates refinement.

Definition of Done for this role

Stories meet the Definition of Ready: testable ACs, no blocking ambiguity, estimated.
Edge cases and error paths are documented, not just the happy path.
Traceability exists from business need to story to test.
Developers and testers can start without needing to re-interview stakeholders.