Gus Dantas
Data services for government: engineering, analysis, evaluation, reporting and automation
0481 870 222
Agencies hold data that answers real questions, but much of it still reaches people as PDFs, spreadsheets and static reports. I clean, migrate and connect that data, analyse and evaluate what it says, report it in dashboards people use and automate the steps your team repeats, with every number traced back to its source. Fixed price, one to six weeks.
I'm a Sydney data analyst and builder with a decade of data work, the last eight years in Australian analytics. My reporting systems run in production today: one assembles 73 of 74 campaign reports in a single batch run.
Core competencies
| Category | What you get | Done before |
|---|---|---|
| M05 Information management | Interactive dashboards in Power BI, Looker Studio or D3, built to WCAG 2.1 AA. Published PDF and Excel statistics rebuilt as dashboards that work on a phone. | An operator dashboard with benchmarks, alerts and drill-downs, used daily by a campaign team. |
| M02 Custom application development | Pipelines that pull, clean, reconcile and load data on a schedule, in Python, SQL and TypeScript, plus Power Apps, Power Automate and SharePoint workflows for approvals, forms and scheduled emails. | A pipeline that loads six ad platforms into one warehouse and drafts the reports. |
| M01 Business analysis | The questions, KPI definitions and data sources agreed in writing before anything is built. | The segmentation framework every supplier incrementality report is built on. |
Tools: Python, R, SQL (BigQuery, Snowflake, Postgres), Power BI, Power Automate, Power Apps, SharePoint, Tableau, Looker Studio, D3, GA4, Git.
Differentiators
- The person you brief builds it, and explains it. One senior analyst from scoping to handover. The walkthrough comes from the person who made every chart.
- Every number traces to its source. Figures are reconciled to the source tables before release, and each chart says where its numbers came from. AI speeds up the build; I check every figure myself.
- A working version in week one. You review a live dashboard every week, so a wrong assumption costs days instead of the whole budget.
- Yours to keep. Built in your own tenant or cloud account, accessible to WCAG 2.1 AA, and handed over with the code and a runbook.
- Priced up front. Day rates and fixed-price packages are on page 2.
Risk management
- Agency data stays in the environment the agency specifies.
- No agency data goes into a third-party AI service without written approval.
- All code and configuration is version-controlled and handed over.
- Continuity through documentation: your staff can run every deliverable from the runbook.
Past performance
| Period | Organisation | What I built | Measured result |
|---|---|---|---|
| 2025 to now | Retail media network of a national specialty retailer, Data & Analytics Lead | Reporting pipeline across six ad platforms, an operator dashboard, six AI data integrations | 73 of 74 campaign reports in one batch run |
| 2022 to 2025 | The Media Precinct, Head of Analytics (Cerebral Palsy Alliance, Dementia Australia, RFDS, Fella Hamilton) | Creative performance dashboard, AI-assisted survey coding, marketing mix modelling | The agency's most requested deliverable; 7,400+ responses coded |
| 2020 to 2022 | AKQA Media (WPP), Senior Data Analyst | Forecast and scenario tool for Afterpay, clustering for Officeworks, segmentation and a refinancing forecast for Aussie Home Loans | Four abandoner personas, each with a recovery lever |
| 2018 to 2020 | Switched On, Data Analyst | Production dashboards and client reporting in Tableau and SQL | Used by the media team every week |
Roles held: Data & Analytics Lead, Head of Analytics, Senior Data Analyst and Data Analyst at the organisations shown.
Packages and prices
| Package | What you get | Time | Price (A$) |
|---|---|---|---|
| Free prototype | A working, unofficial prototype built from your published data, before you commit. Private, no agency branding, taken down on request. | Before you buy | No charge |
| Discovery and scope | Questions, data sources, risks and acceptance criteria in writing. Credited against the build. | 1 week | 6,000 |
| Accessibility check and fix | An existing dashboard tested against WCAG 2.1 AA, then fixed. | 1 to 2 weeks | 5,000 to 9,000 |
| Single dashboard | One data source, one dashboard, refreshed on schedule. | 1 to 2 weeks | 8,000 to 10,000 |
| Statistics to dashboard | PDFs and spreadsheets rebuilt as an interactive, accessible dashboard, with sources on every chart. | 2 to 3 weeks | 12,000 to 18,000 |
| Analysis, evaluation or survey study | A forecast, segmentation, campaign or service evaluation, or survey analysis, with a method note. | 3 to 5 weeks | 15,000 to 30,000 |
| Reporting pipeline | Data pulled, cleaned, reconciled and loaded on a schedule, feeding a live dashboard. Priced by number of sources. | 4 to 6 weeks | 28,000 to 45,000 |
| Run and support | Monitoring, fixes, and changes up to 1 or 2 days a month. Reply within one business day. | Monthly | 1,500 to 3,000 |
Each package is a defined scope with data assumptions, a review point and written acceptance criteria. Work outside the packages: A$1,600 a day for scoping and design (NSW Senior Manager), A$1,200 a day for the build (NSW Senior Consultant), per 8-hour day. Prices exclude GST.
How an engagement runs
1. Discovery, about 1 week, its own package: read the current reports and data. 2. Scope, a few days: deliverables, sources, risks and acceptance criteria in writing. 3. Build, with a working version every week, and 4. Handover, a runbook and a session with whoever maintains it: 1 to 6 weeks together.
Business details
- Name
- Gustavo Dantas Ferreira, trading as Gus Dantas
- ABN
- 20 611 702 297
- Location
- Elizabeth Bay NSW 2011
- Education
- Master of Applied IT, Victoria University. Bachelor of Mechatronics Engineering, UFSJ
- Web
- ghostavo.com/government
- gus@ghostavo.com
- Phone
- 0481 870 222