Data & Operations Analyst

About Sophiie AI

Sophiie AI is an AI-driven office manager for SMB services businesses — built to replace the admin, coordination, and operational overhead that bogs down growing teams. We went from $0 to 7-figure ARR in under 12 months, backed by top VCs, with a lean global team spanning Australia, offshore.

We are genuinely AI-native — not in the "we use an AI tool to summarise meeting notes" sense. We use Claude Code and Codex daily and we’ve built a fleet of specialised AI agents ("The Interns") that handle business operations for us daily: revenue analysis, customer health monitoring, financial reporting, HR scheduling. You’ll be working alongside them — querying them, validating them, and eventually extending them.

About the Role

An analyst role for someone who likes business questions more than code — but is willing to learn to build. This is not a graduate program. It’s not a support role. It’s not a "you’ll get there eventually" job. You’ll get exposure in 18 months that most analysts wait five years for — across data, operations, finance, and AI. The trade is pace and ambiguity.

You’ll operate at the intersection of data, operations, AI, and commercial decision-making from day one. You’ll be exposed to things most people your experience level don’t get near: real-time business analytics, AI agent infrastructure, investor reporting, RevOps, and — over time — building internal tooling with AI that actually ships.

It will be intense. The pace is fast, the expectations are high, and "figure it out" is not a phrase we use sarcastically. But if you’re the kind of person who finds that energising — read on.

Responsibilities

Weeks 1–4: Own the data layer

• Building out our data warehouse insights against an existing spec (dashboards, funnels, cohort views, revenue metrics)
• Maintaining live dashboards the leadership team uses to make weekly decisions
• CRM reporting — pipeline health, MRR, ICP grades, lifecycle funnel
• Getting up to speed with The Interns — learning to query them, validate their outputs, and understand what they can and can’t do

Months 2–3: Go deeper and start touching ops

• Own the weekly dashboard review cycle — catch anomalies, triage issues, escalate what needs engineering
• Support investor data pulls and narrative drafting (you’ll see how leadership packages the business story)
• CRM data hygiene — deduplication, list audits, workflow quality checks
• Start taking on operational work: procurement, contracts, employee onboarding

Month 4+: Build, automate, expand

• As you go deeper, you’ll start extending our internal tools — building small automations and MCP integrations using Claude Code and AI coding tools.
• You’ll also pick up: RevOps workflow monitoring and first-line troubleshooting, Governance and compliance administration.

What This Role Is Not

• It's not an engineering role. If you want to build product, this isn't it. The build work here serves the analysis — not the other way around.
• It's not a 9-to-5. Some weeks will be quiet. Others won't be.
• It's not heavily structured. You'll need to manage your own time, chase the right information, and prioritise without always being told how.
• It's not a "learn first, contribute later" role. The expectation is that you're adding value in week one.
• It's not a pure data role or a pure ops role. You will context-switch constantly — that's a feature, not a bug.

If any of that sounds like a dealbreaker, this probably isn't the right fit. If it sounds like exactly what you've been looking for, keep reading.

If you've read this far and something in this JD made you want to go and answer a business question — or build a small thing to answer it — that's probably a sign.

What You'll Learn

• If you're early in your career and join this role, in 12–18 months you will have:
• Seen the full operational and financial inner workings of a fast-growing SaaS company
• Run a live AI workforce and understood its limits (which most people who "work in AI" never actually touch)
• Written investor reports, managed vendor relationships, owned RevOps data
• Built real things with AI tools — not demos, not proofs of concept, things people use daily
• Worked across every function — finance, engineering, sales, CS, marketing, HR — in a way that's impossible at a larger company

We're not promising it'll be easy. We're promising it'll be worth it.

What We're Looking For

Non-negotiables

• 1–3 years in ops, analytics, or a generalist role at a startup or SaaS company — or a high-signal background that shows you're commercially sharp and self-directed
• Basic understanding of SaaS metrics (such as ARR, churn, LTV, CAC, NRR)
• Working SQL. You can write a query to answer a real business question — not memorised syntax, but you know what JOIN, GROUP BY, and a window function are for.
• You've taken a messy, ambiguous business question, pulled the data, and written up the answer in prose a non-technical leader could read. And you enjoy that work.
• Comfortable in Excel/Sheets and at least one BI or CRM tool
• Strong written communication — you'll write reports, process docs, and messages to the whole team
• Curiosity about AI coding tools. You don't need to have shipped a production system, but you've at least tried Claude Code, Cursor, or similar — and you're ready to learn fast on the job
• Melbourne-based (or willing to relocate)

Great to have

• A built thing using Claude Code or another AI coding tool — an automation, a script, a small internal tool
• CRM experience or certification
• Product analytics or data integration tools
• Australian compliance familiarity (ATO, ASIC, R&D tax incentive)
• Experience working across timezones with distributed teams

Pay range and compensation package

Salary: A$80,000 + super. ESOP participation available for the right candidate.

How to Apply

Send your application to ***email_hidden*** with one of the following 2-min Looms:

• Option A — A business question you investigated. Walk us through how you scoped it, pulled the data, what you found, and how you communicated the answer. Use of AI tools encouraged but not required. We want to see how you think about ambiguity, evidence, and explanation.

• Option B — A thing you built with an AI coding tool. Claude Code, Cursor, or similar. An automation, a script, a dashboard, a small internal tool — what it does, why you built it, how.

Either is fine. Pick the one that shows you at your best.

And your LinkedIn URL. Nothing else. No CV, no cover letter. If your Loom clears the bar, we'll reach out within a few days.