Business Analyst - Agentic Engineering

Nimbly; Sydney; On-site; AUD 90,000 - 140,000; Full time

Job summary

Nimbly is seeking an experienced business analyst to partner with customers in Sydney. You will translate messy operations into AI-ready specifications that our engineering team can build against.

You’ll run customer discovery, define target-state workflows, specify data needs, capture tacit knowledge, and write precise specs while owning the end-to-end delivery and maintaining strong stakeholder relationships.

Qualifications

• 3+ years in business analysis or similar customer-facing roles.
• Strong BPMN/process modelling skills and ability to switch between narrative and detail.
• Hands-on experience documenting integrations and data contracts (APIs, events, master data).
• Confident presenting to operators and executives; can run workshops and steering meetings.
• Interest in AI agents and experience with LLMs or agentic systems is a plus.
• Clear, concise written communication suitable for engineers and sign-off by executives.

Responsibilities

• Run discovery with customers: lead workshops, shadow operators, interview SMEs, and map current-state processes.
• Define target-state workflow: decide ownership, handoffs, exceptions, and done criteria.
• Specify data dependencies: identify entities, data quality, and master records.
• Capture institutional knowledge: surface rules and heuristics for AI tooling.
• Write detailed specifications: triggers, inputs, decisions, tool calls, outputs, and success criteria.
• Own the customer relationship through delivery: validate prototypes, perform UAT, adjust specs.
• Create reusable patterns: collaborate with Product to productise where appropriate.

Job description

Work directly with customers to document the target-state workflows, data and products. You translate messy, real-world operations into clean, AI-ready specifications that our engineering team can build against.

What you'll do

• Run discovery with the customer. Lead workshops, shadow operators, interview SMEs and reverse-engineer business processes. Produce the current-state map quickly and honestly.
• Define the target-state workflow. Re-design each in-scope process for an agent-first world. Decide what the agent owns, what the human owns, where the handoffs are, what the exception paths look like, and what "done" means
• Specify the data. Identify the entities the workflow depends on (customer, job, asset, invoice, technician), where they live today, how clean they are, and what the master record needs to look like. Flag gaps before engineering hits them
• Capture institutional knowledge. Surface the tacit rules and tribal heuristics that experienced staff carry in their heads — pricing exceptions, supplier quirks, regional rules, "never do X on a Friday" - and structure them for AI tooling
• Write detailed specifications. Produce the artefact engineering builds against: triggers, inputs, decisions, tool calls, outputs, edge cases, success criteria, and the human checkpoints. Clear enough for an engineer; precise enough that the customer can sign it off
• Own the customer relationship through delivery. Validate prototypes with the customer, run UAT, manage scope conversations honestly, and adjust the spec when reality pushes back
• Patterns you find on one engagement become reusable modules and reference workflows. You will work with Product to harden what should be productised and keep what stays bespoke clearly bespoke

What we're looking for

• 3+ years in business analysis, solution consulting, process design, or similar customer-facing analyst roles
• Strong workflow and process modelling skills - BPMN, swimlanes, or equivalent; comfort moving between high-level narrative and step-level detail
• Hands-on experience documenting integrations and data contracts (APIs, events, reference data, master data)
• Confident in front of both operational staff and senior stakeholders - you can run a workshop and a steering meeting
• Genuine interest in AI agents and how they change the shape of business workflows; prior exposure to LLMs or agentic systems is preferred
• Direct, concise written communication - your specs are read by engineers and signed off by executives