Ange-Ghislain KOUAME · Data Engineer in Abidjan

§ 01 / 08

Two crafts, one standard.

Data for those who decide, design for those who need to be seen. The same care in both: clear, tested, delivered on time.

Pole 01 · 04 offers

Data

Make your figures reliable, documented and useful for decisions.

Custom data pipelinesTurn scattered Excel extracts, exports and databases into a clean, documented warehouse that refreshes without surprises.
Decision dashboards with Power BIDashboards that answer your management questions, with figures you can trace back and defend in a meeting.
Data documentation to DDI standards and NADA cataloguesDocument your surveys and administrative datasets to international standards, and publish them in a searchable catalogue with access control.
Data quality audit and anomaly detectionMeasure how reliable your data really is, and find the records that deserve a closer look before decisions are made.

Pole 02 · 04 offers

Studio

Make your activity visible, recognisable and easy to contact.

§ 02 / 08

See the work in motion.

Two films made for this site, rendered frame by frame with original music. They start muted; turn the sound on if you wish.

Write to me See the offers

§ 03 / 08

A report you can explore.

Pipeline, audit, documentation and dashboard applied end to end to a public dataset: you judge the work itself, not a promise.

Live demonstration · public data

French property prices, 2025

3.8 million public lines handled as in a client mission: controlled pipeline, quality audit, DDI documentation and a six-page report in the Power BI manner, with filters, cross-filtering and a detail page for each department.

Open the report See the method

Dwelling sales retained: 732,164
Flats, median price per m²: €3,370/m²
Houses, median price per m²: €2,200/m²

§ 04 / 08

A studio, pieces you can touch.

Brand, print, web and motion: everything here was designed for my own brand, the way I would design yours.

Brand identity

Mark, palette and typefaces: the guide behind every page of this site.

AK Data quality audit Your figures deserve to be believed.

Completeness, duplicates, inconsistencies, with the rows concerned Unusual records flagged, with criteria set in advance Recommendations to fix the data at the source Let's talk about your data Ange-Ghislain KOUAME · angekouame.vercel.app

Flyer

A5 format, written to sell one offer in five seconds. The QR code leads to the offer.

Business card

Front, back and a QR code to the site. Click or tap to turn it over.

732,164 sales retained

Dwelling sales per month, France, 2025

Animated figure

A key figure set in motion, here from the property prices demonstration.

Home page of this portfolio on a computer. — Showcase website

This portfolio: bilingual, accessible, readable on a phone, with its own administration.

The same page on a phone. — Showcase website

This portfolio: bilingual, accessible, readable on a phone, with its own administration.

Talk about a studio project See the studio offers

§ 05 / 08

Selected engineering work.

Case studies written as engineering documents: problem, architecture, decisions, limits.

In internal use
Public Procurement Data Platform

From a raw spreadsheet extract to certified, traceable and documented data: layered ETL, rule catalogue, star schema, semantic model and Merkle integrity certificates.

Period

2026, ongoing

Context

Ivorian Ministry of Economy, Finance and Budget · internal tool

Stack

Python · pandas · Jupyter · SQL Server 2022 · T-SQL

Prototype
Explainable Risk Scoring and Unsupervised Anomaly Detection

An auditable scoring grid, forensic tests and a Keras autoencoder benchmarked against FAMD, with success criteria fixed before computation and limits reported as found.

Period

Oct 2026

Context

Analytical extension of the procurement data platform · awaiting business validation

Stack

Python · pandas · NumPy · scikit-learn · SciPy

In development
Secured Data Catalogue and Internal Portal

A DDI/NADA metadata catalogue hardened for restricted microdata, a tested library module, and server upgrades rehearsed on a full replica with scripted rollback.

Period

2026, ongoing

Context

Internal portal of the Financial Control Directorate · not exposed to the internet

Stack

NADA 5.4 · World Bank Metadata Editor · PHP 8.4 · MySQL 8.4 · Apache 2.4

Delivered
Budget Execution Audit Analytics

An end-to-end audit pipeline on official budget data for a ministry section, from raw spreadsheets to an interactive diagnostic and an automated audit narrative.

Period

Mar 2026

Context

Ministry of Agriculture and Rural Development · budget section 229, fiscal year 2024 · official State data

Stack

Python · pandas · Metadata Editor · Power Query · Power BI

All case studies

§ 06 / 08

Career, in brief.

Since 2020 I have worked at the Direction du Contrôle Financier (Financial Control Directorate) of the Ivorian Ministry of Economy, Finance and Budget, first administering the electronic document management system, then as a data analyst and data engineer.

My current focus is the industrialisation of public procurement data: an ETL pipeline with a rule catalogue and quarantine, a star-schema warehouse, a traceable Power BI model, integrity certificates, and DDI metadata published to an access-controlled catalogue.

I work best with organisations that treat data platforms as products: international organisations, financial institutions and data teams that value reliability, governance and measurable impact.

2017
Networks and software engineering studies
2020
Information systems, Financial Control Directorate
2022
Data analyst, then data engineer
2026
Public procurement data platform

Experience

Role: Data Engineer, Financial Control Directorate, Ministry of Economy, Finance and Budget
Now building: Certified, traceable public procurement data platform
Core stack: Python · SQL Server · Power BI · DDI
Experience: 6 years in public financial information systems
Languages: French (fluent), English (professional use)
Open to: Freelance engagements, consulting, collaborations

§ 07 / 08

Good questions, straight answers.

What recruiters and partners usually ask first.

Are the professional figures real?

Yes. They come from the systems described in the case studies and are published with the authorisation of the Financial Control Directorate. Detailed audit findings and personal data stay internal.

Can I see the code?

The production repositories belong to the institution and stay private. The team repository of the retail sales datamart is public on GitHub, and the case studies describe how each system is built.

All case studies

What do the status labels mean?

Every project shows where it really stands. The shape of the marker repeats the label, so colour is never the only cue.

In internal use: Running in the employer's day-to-day work, without public access.
In development: Under construction.
Prototype: Working proof of concept, not in production.
Delivered: Completed and handed over.
Academic: Coursework or training project.
Demonstration: Reproducible work on public data, built to show a method.

How can I check the certifications?

On the About page, every certificate that has a public verification page links to it, next to its credential ID.

Certifications

Where are you based, and how do you work?

In Abidjan, Côte d'Ivoire. I work in French, use English professionally, and am available as a freelancer for consulting and delivery engagements, remote or on site.

Contact

§ 08 / 08

Let's talk about your data platform.

Available as a freelancer for consulting and delivery engagements, and for collaborations with public institutions and development organisations.

angeghisk@gmail.com

Email me LinkedIn Download CV (PDF) Contact

Engineering reliable data systems. Enabling better decisions.

Two crafts, one standard.

Data

Studio

See the work in motion.

Where to start

You have a project

You want proof

You want to know who I am