Data Analyst · Trade & maritime data

I turn messy data into dashboards teams can trust.

I build Power BI reporting across 30+ sources and Python-to-Snowflake pipelines for 40+ countries. My tools cut product onboarding from about 30 days to 3–4. I now lead a team of four at DNEXT Intelligence.

Open to Data Analyst roles and freelance projects · Based in Tunisia, open to remote, on-site or relocation

Skander Ghamgui

Skander Ghamgui
Data Analyst and team lead at DNEXT Intelligence

0130 → 3–4 days

days to onboard a product

My automation tool replaced a month-long setup queue.

0270%+

fewer manual follow-ups

Arrival prediction filled the gaps when ports reported late.

0340+

countries in one daily pipeline

Emailed PDF and Excel reports turned into clean daily data.

0420+

data sources automated

Up to 80% less processing time on customs and statistical data.

054

team members led

Promoted from Junior Data Analyst to team lead in 10 months.

Projects

What I built, and the time it saved.

Real projects from my work in trade and maritime data. Each one shows the problem, what I built, and the result.

01Market analysis · Commodity trade
Illustrative port scene representing commodity trade analysis
20–30%of trade actually covered, on products the client thought it led

How much of the market are we missing?

An analytics tool that compares vessel lineup volumes with official customs figures, so a commodity inspection company could see its real market coverage and who to approach next.

02Data platform · Maritime intelligenceLIVE DASHBOARD
Synthetic lineups dashboard preview with summary metrics and vessel records
40+countries, one reliable flow

40+ countries of shipping reports, turned into one daily dataset.

An automated operating system that turns inconsistent maritime reports into clean, daily, decision-ready datasets across more than 40 countries.

03Business intelligence · Decision support
Analyst working with a data spreadsheet on a laptop
30+sources reconciled

30+ trade sources that disagreed, compared on one dashboard.

A Power BI intelligence layer that reconciled trade and maritime sources, made discrepancies visible and gave senior analysts a shared global view.

04Automation · Internal product tooling
Team collaborating around a laptop in an office
30 → 3-4days to onboard

A month-long onboarding process, cut to 3–4 days.

A desktop workflow that coordinated APIs, HS-code configuration, metadata validation and scheduling so new products and countries could move faster.

05Data quality · Entity resolution
Blue-lit server racks representing reliable data infrastructure
Auditableevery merge reviewable

Finding the same ship under different names.

A production-grade quality layer combining deterministic cleanup, fuzzy matching and maritime timing heuristics to resolve noisy vessel records.

06Web scraping · Data engineering
Laptop displaying source code in a developer workspace
20+sources automated

20+ public data sources, collected automatically.

More than 20 source-specific collectors that transformed statistical and customs websites into repeatable, multi-year ingestion workflows.

07Freelance · Jul 2026 – Aug 2026 · E-commerce data engineering
Modern furnished living room representing an e-commerce catalog
28.6Kproduct records processed

Eight product catalogs merged into one online store.

A real, anonymized freelance engagement that transformed 28,646 source product records into reviewed, validated and WooCommerce-ready catalog workflows.

Capabilities

The whole data chain, from source to dashboard.

My advantage is range with continuity: I can trace a problem from source acquisition to modeling, monitoring and the final decision surface.

COLLECT → CLEAN → MODEL → EXPLAIN → IMPROVE

How I work

A reliable system starts with the right questions.

I do not automate ambiguity faster. I make the operation understandable, separate the moving parts and leave the team with a system it can trust and operate.

01

Start with the operation

I map who uses the data, where it breaks and which decision the finished system must improve before choosing the technology.

CONTEXT BEFORE CODE
02

Build in clear layers

Collection, normalization, business rules, quality controls and delivery stay separated so the system remains understandable and adaptable.

MODULAR BY DESIGN
03

Make quality visible

I prefer explicit validation, exception queues and audit trails to pipelines that look successful while silently losing trust.

FAIL LOUDLY · REVIEW CLEARLY
04

Leave an operating system

Documentation, ownership, monitoring and coaching are part of delivery so a team can run and improve the work after handoff.

BUILT TO BE OPERATED

Experience

Hands-on builder. Accountable leader.

My trajectory moved quickly from analyst to system owner and team lead, without losing the technical work that makes delivery credible.

01

Jul 2026 - Aug 2026

Remote

Freelance

Freelance Data Analyst

Scraped product catalogs from eight websites (28,646 records) and loaded them into a client's WooCommerce store, with checks before every import.

  • Web scraping
  • Data quality
  • Python
02

May 2025 - Present

Sousse, Tunisia

DNEXT Intelligence SA

Head of Lineups Team

Leading a four-person team responsible for global maritime lineups and agricultural trade-flow data, while remaining hands-on across architecture, automation and quality. I built the data quality tools the team uses today, from defining the rules to running them.

  • Global pipeline ownership
  • Data quality tools
  • KPIs, SLAs and coaching
  • Cross-functional product delivery
  • Weekly executive reporting
03

Jul 2024 - Apr 2025

Sousse, Tunisia

DNEXT Intelligence SA

Junior Data Analyst

Built source collectors, ingestion workflows, Snowflake queries and multi-source BI views for international trade datasets.

  • 20+ automated scrapers
  • Power BI comparisons
  • Snowflake production work
  • Docker standardization
04

Jan 2024 - Jul 2024

Tunis, Tunisia

Le Sportif

R&D Engineer

Built Odoo inventory and reporting modules that reduced stockouts by 50%. Migrated ERP data and connected e-commerce operations to reduce manual synchronisation.

  • 50% fewer stockouts
  • ERP migration
  • Odoo modules
  • PrestaShop integration
  • Inventory reporting
05

Feb 2023 - Aug 2023

Tunis, Tunisia

Sagemcom

Business Analyst Intern

Automated recurring reporting and transformed PDF data into structured JSON for downstream smart-meter workflows.

  • Reporting automation
  • PDF to JSON
  • Data preparation
  • Process improvement

EDUCATION

Master's in Business AnalyticsESPRIT School of Business · 2021-2023
Bachelor's in Management ScienceESPRIT School of Business · 2018-2021

Services

How I can help, in your team or on a project.

I work best on bounded, high-leverage problems where a reliable data flow can remove recurring effort or unlock better decisions.

01

Scraping & source extraction

Turn websites, portals, PDFs and spreadsheets into structured, repeatable datasets.

  • Source assessment
  • Resilient collector
  • Clean schema
  • Run documentation
↗
02

Workflow automation

Replace fragile copy-paste operations with validated flows and clear exception handling.

  • Process map
  • Automation build
  • Quality controls
  • Operational handoff
↗
03

Power BI decision systems

Design models and dashboards that expose the signal, the context and the reliability of the data.

  • Semantic model
  • DAX measures
  • Executive views
  • Refresh monitoring
↗
04

SQL & data quality

Make production data faster, more consistent and easier to trust across the analytical workflow.

  • Query optimization
  • Validation rules
  • Reconciliation
  • Audit-ready outputs
↗

Contact

Hiring a Data Analyst, or need one for a project?

Email me, download my CV, or tell me about your project below.

Available for select freelance projects and international opportunities.

Ask Skander

SKANDER.KB / APPROVED KNOWLEDGELOCAL · PRIVATE

ASK SKANDER

Skip the scrolling.
Ask what matters.

Answers come only from the experience and projects published on this site.

ANSWER / IDENTITY

I am strongest where the data is operationally important but difficult to trust: mixed files, shifting source formats, duplicate entities, fragmented manual steps and dashboards that need better foundations. I turn that ambiguity into a system people can run and act on.

See my services