AI First Business Analyst

PhoenixDX is a digital transformation and software delivery partner helping organisations achieve meaningful business outcomes through modern platforms, structured delivery and trusted advisory capability. We combine consulting, architecture and engineering excellence to move customers from idea to measurable value with clarity and confidence.

This role sits in our AI-First Innovation Hub — squads where agents carry the delivery volume and people direct, review and own the calls that matter. The toolchain itself is AI-powered: an agentic code-generation pipeline, AI-assisted modernisation tooling and AI-native product features, all on a single Source of Truth with human gates wherever work touches production.

What you will do...

The AI First Business Analyst owns the problem space from the first stakeholder conversation to delivery acceptance — a combined discipline that compresses business analysis, UX and test-writing into one role, operating with an orchestrated AI toolkit.

WAS writing specs by hand from scattered docs  →  NOW sharpening intent, edge cases and business rules

This is not a traditional BA with AI tools bolted on. It is a combined discipline — analysis, design and test craft in one — that enables a level of throughput and quality no traditional analyst structure can match.

Requirements & Discovery

• Lead stakeholder workshops, pain-point analysis and requirements mapping on complex enterprise engagements

• Draft BRDs, process journeys and traceability artefacts in hours, using AI-assisted analysis to surface gaps, implicit requirements and hidden dependencies

• Maintain requirements integrity across long build cycles — traceability from business intent to acceptance

Design & Prototyping

• Build high-fidelity UI in Figma directly from requirements

• Generate clickable prototypes for rapid stakeholder validation before a line of code is written

• Keep requirements, architecture and design in continuous sync

Quality & Acceptance

• Create test scripts from acceptance criteria as a natural output of story craft, not a downstream workstream

• Flag coverage gaps, missing acceptance criteria and untested integration boundaries before sprint execution

• Support UAT and validate delivered functionality against intent

AI-Augmented Analysis

• Run multi-pass requirements analysis, gap detection and artefact generation with an orchestrated AI toolkit

• Deploy browser-based test agents to assess application state and triage defects

• Know when AI output is right, plausibly wrong or confidently wrong — and never ship a draft unreviewed

Delivery Flow & Backlog

• Lead AI-assisted backlog refinement: propose refinements, cluster defects, validate with the Product Owner, act

• Produce development-ready stories — design, acceptance criteria and test scripts — in one integrated workflow

• Drive sprint precision through AI-assisted dependency mapping and scope-risk detection

Domain & Workflow Depth

• Build deep understanding of the business domain and its workflows — not just the ticket in front of you

• Represent business intent accurately in architecture and solution conversations, and challenge decisions that create downstream requirements risk

• Apply software design principles (separation of concerns, API contracts, event-driven design) to spot delivery friction before it becomes a defect

Who you are...

Core — Required

• 5+ years as a Business Analyst or Technical Analyst on complex, enterprise custom software delivery — not SaaS configuration or light-touch product work

• Held requirements integrity on systems that matter, across both business and technical stakeholders

• Familiarity with OutSystems or a comparable low-code rapid-application-development platform

• Sufficient technical depth to participate in architecture and solution design — design patterns, integration principles and data modelling

Core — Preferred

• Consulting or professional-services background

• Mentoring or upskilling analysts in AI-augmented practice

• Agile scaling frameworks (SAFe, Scrum or similar)

• Tricentis, Lovable or agentic browser-testing tools

AI-First Practice — Required

• An AI-integrated practice you have shaped deliberately — you can say what you delegate to AI, what you share with it, and where you keep full human ownership, and why.

• Discriminating judgement about AI output — you know when a result is right, plausibly wrong, or confidently wrong, and you do not ship AI drafts without review.

• You stay current on the AI tooling ecosystem (agent platforms, Claude skills/plugins, low-code AI extensions), trial new capability against real delivery problems, and adopt what genuinely raises the bar.

AI-First Practice — Preferred

• An AI-integrated practice you have shaped deliberately, with clear delegate / share / own boundaries

• Daily work with the tooling stack: LLM agents, Jira / Rovo, Figma and test-automati