Data analyst, humanitarian data
Turn programme, needs-assessment and funding data into analysis that NGO and UN decision-makers can act on. Remote, per project, paid on delivery.
- কাজের ধরন
- Freelance, per project
- অবস্থান
- Remote
- সময়ের মিল
- At least 3 hours with Central European Time
- প্রকাশিত
- 15 অক্টোবর, 2026
- ভাষা
- English (working language) · Ukrainian, Spanish, French or Arabic are a plus
- দক্ষতা
- SQL (PostgreSQL)Python or RMetabase, Power BI or similarKoboToolbox and survey dataHumanitarian indicators and reporting cycles
About the work
baena.ai builds data tools and analysis for humanitarian organisations: dashboards on pooled-fund allocations, labour-market analysis of the aid sector, cash-assistance reporting, needs-assessment pipelines. The work arrives as scoped engagements with a client, a deadline and a deliverable. We are looking for analysts who can take one of those engagements from a raw export to a finished product with light supervision.
What you will do
- Clean, model and document datasets from KoboToolbox, ActivityInfo, OCHA FTS, ReliefWeb and clients' own spreadsheets.
- Write the SQL that feeds a dashboard or a report, and keep it readable for the next person.
- Build Metabase dashboards and short written analyses that answer the client's question rather than describe the data.
- Check your own numbers against the source and flag what does not add up before the client does.
- Hand over through the client portal: files, notes and a short description of what was done and why.
What we look for
- Two or more years working with programme, monitoring or funding data in the humanitarian or development sector, or equivalent analytical work with messy survey data.
- Confident SQL on PostgreSQL, and Python or R for anything SQL cannot do cleanly.
- You have shipped at least one dashboard or recurring report that people actually used.
- Clear written English. Most of the work is explained in writing, inside the portal.
- Comfortable with the sector's vocabulary: clusters, pooled funds, MPCA, 3W/4W, indicators and targets.
Nice to have
- Geospatial work (PostGIS, QGIS) or experience with mapping deliverables.
- Experience with n8n or another workflow tool for scheduled pipelines.
- A second working language used in a response context.
How the engagement works
- Engaged per project, remotely, on a written scope with deliverables and a fee. There is no employment contract and no retainer by default.
- Paid on delivery of each agreed milestone, by invoice.
- You are introduced to the client by name and work through the same portal and under the same confidentiality terms as the practice itself.
- Client data stays on EU-hosted tools under our control; sensitive material is processed on local AI models, never on third-party AI services.
- Everything that reaches a client is reviewed by a person. AI can speed up the work; it does not sign it off.
How to apply
Use the form on this page: upload your CV as a PDF and, if you like, a cover letter. The letter is the place for one or two links to work you can show (a dashboard, a report, a notebook, a repository), your availability over the next three months and an indicative day rate.