Senior Data Engineer

COMPENSATION — PLEASE READ FIRST

This is an unpaid, part-time pre-launch opportunity. Properta cannot offer a salary or stipend at this stage. We are looking for a senior engineer who can commit reliable hours alongside existing work or other commitments.

If Properta launches successfully and generates sufficient revenue or secures investment, we hope to explore a paid, full-time role with the right person. A future role, salary and start date are not guaranteed.

Before work begins, we will agree in writing on your weekly availability, an achievable initial deliverable and a review date. Please apply only if these terms are workable for you.

ABOUT PROPERTA

Properta AI Technologies Ltd is building an independent property intelligence platform. Our first product helps people assess residential properties in Dubai using property data and clearly explained evidence.

A Properta report should show a supported price range, the comparable transactions behind it, the strength and limitations of that evidence, and when the available data is insufficient to support a conclusion. Rental and return information should appear only when the underlying evidence and cost assumptions support it.

We already have substantial property datasets, analytical components and a customer-facing experience. We now need to make the data foundation dependable, traceable and ready for production.

THE ROLE

You will report to Properta’s CTO and work closely with the backend, product and analytical teams. You will take hands-on ownership of agreed data engineering deliverables: from understanding source files and defining transformations to implementing pipelines, testing their outputs, deploying them and investigating failures.

Our first production priority is analysis of ready residential property sales. Rental and return analysis is a later, conditional phase. Off-plan analysis has separate data and evidence requirements. Your initial scope will be chosen to fit the hours we agree; you will not be expected to complete every item below at once.

WHAT YOU WILL WORK ON

• Data ingestion and refresh: Build reliable processes to ingest large, authorised property datasets into Azure-based storage and analytical systems. Handle recurring deliveries, full-file replacements or incremental updates as appropriate. Make refreshes repeatable and safe to retry.

• Data modelling: Turn raw property, building, project, transaction and, where relevant, tenancy records into documented datasets that services and analysts can use. Define keys, relationships, data types and clear rules for records that cannot be reliably linked.

• Cleaning and matching: Investigate missing values, inconsistent identifiers, duplicate records, unusual dates, implausible areas and prices, and changes in source data. Implement defensible normalisation and matching rules without presenting uncertain matches as facts.

• Property and comparable data: Prepare the data needed to identify a subject property and select relevant comparable transactions. Work with the CTO on rules for location, property type, bedrooms, size, transaction date, exclusions and minimum evidence. Keep ready and off-plan evidence separate.

• Analytical implementation: Implement and test agreed calculations, data filters, price-range inputs and confidence measures. Surface insufficient or conflicting evidence so the report service can respond appropriately. You will help improve the rules through testing and evidence; product and analytical decisions will be agreed with the CTO.

• Provenance and reproducibility: Preserve where each record came from, when it was received, how it was transformed and which rule or dataset version contributed to an output. Make it possible to investigate a customer-facing result and reproduce it against the appropriate data snapshot.

• Data quality and validation: Build automated checks for schema changes, row counts, missing fields, duplicate keys, broken joins, unexpected distributions and stale deliveries. Reconcile pipeline outputs with source files and test known property examples before changes reach production.

• Performance: Design practical data structures and optimise SQL, storage and processing for datasets containing millions of records. Help ensure report requests can retrieve the evidence they need without repeatedly processing the entire source dataset.

• Production operations: Add useful logging, metrics, health checks and alerts for data pipelines. Document failures, recovery steps and safe reprocessing. Work with the Azure and backend engineers on access controls, deployment, backups and reliable delivery of analytical outputs to application services.

WHAT WE ARE LOOKING FOR

• Strong hands-on Python and SQL experience with large datasets, ETL/ELT pipelines, data modelling and database performance.

• Experience taking data pipelines beyond a prototype into a production environment, including testing, monitoring, troubleshooting and support.

• Practical experience with data cleaning, record linkage or entity matching, validation and handling incomplete or conflicting source records.

• Ability to explain the assumptions behind a transformation or analytical rule and show how you checked whether its outputs were correct.

• Experience working with PostgreSQL or another production relational database.

• Familiarity with Git, code review, automated testing and repeatable deployment.

• Clear communication with engineers and product colleagues, including raising data limitations early rather than concealing them.

We are primarily seeking 5+ years of relevant professional experience, or equivalent evidence of substantial hands-on ownership.

Experience with Microsoft Azure, Azure Blob Storage, Azure Data Factory, Databricks, Spark, DuckDB or similar tools is valuable. Experience with real estate, government, financial or other complex administrative datasets is particularly relevant. You do not need to have worked in real estate before.

LOCATION AND AVAILABILITY

This role is fully remote. Applicants must be based in Dubai or India and available for regular overlap with Dubai business hours.

Please tell us how many hours you can reliably contribute each week. We will agree on an initial milestone that fits that commitment.

HOW TO APPLY

Apply with your CV or LinkedIn profile. In your application, please briefly describe:

1. A production data pipeline you personally built or substantially improved.
2. A difficult data-quality or record-matching problem you solved and how you validated the result.
3. Your location, weekly availability and whether the unpaid, part-time terms are workable for you.

Shortlisted applicants will be invited to complete a questionnaire.