Senior Business Analyst (AI)

Introduction

• Turn vague business problems into shaped, sized, buildable AI and automation use cases, and stay with them through delivery until they are live, adopted and measured.

• The role sits at the front of the funnel and the back of it. At the front: process discovery, feasibility, data availability, benefit sizing, and the honest call on whether a use case deserves to be built at all. At the back: requirements, test design, user acceptance testing, cutover and hypercare. In between, it is the connective tissue between business owners who know the process and engineers who know the platform.

• As the estate moves from rules-based automation toward agentic, data-driven solutions, this role shapes that shift at process level, identifying which workflows should be re-architected around agents rather than automated as they stand, and which should not be touched at all.

Duties & Responsibilities

Use case shaping and demand management

• Run discovery with business units: process walkthroughs, pain-point mapping, volume and effort baselining, root-cause analysis.

• Frame the problem before the solution. Establish what the process does today, what it costs, and where value leaks.

• Assess feasibility across data availability and quality, technical fit, process stability and change appetite, and document the verdict.

• Recommend the right intervention: agentic solution, classical machine learning, robotic or workflow automation, straight process improvement, or no build. Be willing to recomment no build.

• Build the business case on quantified, sourced baselines, sizing benefits conservatively and stating assumptions explicitly.

• Maintain the demand pipeline and prepare use cases for prioritisation gates.

Process and data analysis

• Produce as-is and to-be process models (BPMN or equivalent) covering volumes, handoffs, exception paths and decision points.

• Decompose processes into tasks suitable for agent design: inputs, tools, data sources, decision authority, escalation triggers and human-in-the-loop points.

• Trace and validate data sources: lineage, ownership, refresh cadence, quality, completeness and access path. Surface data gaps before build starts, not during it.

• Document business rules, edge cases and exception handling that engineers would otherwise discover late.

Requirements and delivery support

• Write functional and non-functional requirements, user stories and acceptance criteria that engineers can build from without re-interviewing the business.

• Define non-functionals explicitly: latency, throughput, availability, accuracy thresholds, cost per transaction, audit and retention requirements.

• Work inside the delivery team through the sprint cycle: refinement, clarification, demonstration feedback and defect triage.

• Design and run test scenarios including edge and failure cases, co-ordinate user acceptance testing and drive sign-off.

• Support cutover, hypercare and handover to operations, producing runbooks, standard operating procedures and user documentation.

Value and measurement

• Establish the pre-implementation baseline for every use case, agreed with the business and Finance where relevant.

• Work with data science on measurement design so that benefit claims survive scrutiny.

• Track realised benefit after go-live and report shortfalls early and honestly.

Cloud and platform fluency

• Understand the target architecture well enough to shape solutions realistically: cloud data platforms, integration and API patterns, batch versus real-time processing, model and agent serving.

• Factor consumption cost into solution shaping, including model choice and infrastructure cost per use case.

• Work with Technology on environment, access and integration dependencies, raising them early enough to be resolved.

Governance and risk

• Ensure each use case meets data privacy (POPIA), consent, security, model governance and audit requirements, and assemble the supporting artefacts.

• Document decision authority and controls for agentic use cases: what the agent may decide autonomously, what requires human approval, and what the audit trail must capture.

Team contribution

• Mentor junior and mid-level analysts, and raise the standard of discovery and requirements artefacts across the team.

• Maintain and improve the team’s shaping toolkit: templates, feasibility checklists, benefit sizing method and definition of ready.

Success measures: first 12 months

• Every use case entering build has a documented baseline, a validated data path and an agreed measurement approach.

• Requirements churn and mid-sprint scope discovery reduce measurably.

• Time from demand intake to a decision-ready business case shortens quarter on quarter.

• Use cases reach user acceptance testing with fewer defects traceable to ambiguous or missing requirements.

• Business owners can articulate their own process and its target state unaided.

• Weak use cases are stopped or redesigned at shaping stage rather than at build.

Desired Experience & Qualification

Essential

• Bachelor’s degree in Information Systems, Engineering, Computer Science, Business or a related field.

• 6+ years’ business analysis experience, including 2+ years supporting AI, machine learning, data or intelligent automation delivery.

• Demonstrable record of taking use cases from problem statement to live production and measured benefit, not requirements documentation alone.

• Cloud fluency: hands-on exposure to GCP or AWS data and AI services, with an understanding of integration patterns, APIs and consumption-based cost models.

• Independent SQL capability: able to profile a dataset, validate a source and interrogate a result without waiting for an analyst.

• Working understanding of AI solution patterns: classical machine learning, LLM-based systems, retrieval and grounding, agent orchestration, evaluation and guardrails. Enough depth to shape realistically; build skills not required.

• Process modelling discipline (BPMN or equivalent).

• Agile delivery experience within a product or squad structure.

Preferred

• Business analysis certification (CBAP, CCBA or IIBA equivalent).

• Intelligent automation platform exposure: robotic process automation, intelligent document processing, workflow orchestration, including the judgement to know where these are the wrong answer.

• Process improvement qualification (Lean, Six Sigma) or process mining experience.

• Telecommunications, financial services or large-scale operational business background.

Competencies

• Structured problem decomposition · workshop facilitation and stakeholder interviewing · precise written communication · commercial numeracy · scepticism toward unvalidated assumptions · comfort challenging senior stakeholders on scope and feasibility · persistence through discovery where processes are undocumented.

Kindly regard your application as unsuccessful if you have not heard from the agency within 2 weeks.