Data Analyst · Business Intelligence

Kamel Aggoun

I turn business data into clear analysis, dashboards, and reporting.

Python · SQL · Power BI · Excel

I work end-to-end on data projects: cleaning raw datasets, writing analytical SQL, building Power BI dashboards, and explaining what the numbers mean in plain business language.

View Projects Download CV
Open to Data Analyst & BI roles · Global Khenchela, Algeria

01 · About

About me

Portrait of Kamel Aggoun

I'm a data analyst focused on business analysis and business intelligence. I take a business question, work through the data behind it, and return an answer that is supported by analysis, validated calculations, and reporting someone can act on.

My working toolkit is Python (Pandas, NumPy, Matplotlib, Seaborn, statsmodels), SQL, Microsoft Excel, and Power BI. In practice that means cleaning and preparing data, exploratory analysis, writing analytical SQL, defining KPIs, and building dashboards and reports.

My degree background is in English Language and Culture, and I treat it as a supporting strength rather than my professional identity: it is why I document analysis carefully and explain findings clearly to non-technical readers. My portfolio holds six end-to-end projects spanning HR, customer, forecasting, SQL, financial, and supply-chain analytics.

02 · What I do

What I do

Three things I do in every project, from the first business question to the final report.

Data Analysis

Clean, transform, and explore business datasets to answer a defined question.

  • Data cleaning & preparation
  • Exploratory data analysis
  • KPI definition & validation

Business Intelligence

Build Power BI dashboards and Excel reports that make performance visible.

  • Dashboard design & reporting structure
  • Interactive reports with filters and slicers
  • KPI reporting for stakeholders

Business Analytics

Analyse commercial performance across sales, customers, finance, HR, and operations.

  • Customer segmentation & retention
  • Forecasting & time-series analysis
  • Findings translated into recommendations

03 · Skills

Skills & tools

The technologies and methods I use to collect, clean, analyse, and present data.

Programming

  • Python
  • Pandas
  • NumPy

SQL & Databases

  • SQL
  • SQLite
  • MySQL

Business Intelligence

  • Power BI
  • Tableau

Data Visualization

  • Matplotlib
  • Seaborn
  • Dashboard Development

Statistics & Forecasting

  • statsmodels
  • Time Series Analysis
  • Forecast error measurement (MAPE)

Excel

  • Microsoft Excel
  • Spreadsheet reporting

Analytical Methods

  • Data Cleaning
  • Exploratory Data Analysis
  • Business Analysis
  • KPI Definition
  • Cohort Analysis
  • RFM Segmentation
  • Customer Lifetime Value
  • Pareto Analysis
  • Reporting

How I work

  1. Collect
  2. Clean & prepare
  3. Explore
  4. Analyse in SQL
  5. Validate
  6. Visualise
  7. Report
  8. Recommend

Python · Pandas · SQL · Excel · Power BI

04 · Projects

Selected projects

Six end-to-end analytics projects. Each one starts from a business question and ends with analysis, validation, and a dashboard or report. The three featured projects are the deepest technically; the remaining projects follow the same structure and cover additional business areas.

Featured projects

Strongest technical depth and clearest business value.

Financial performance dashboard showing revenue, cost, profit, and margin KPIs

Financial Performance Analysis

Revenue, cost, profitability, and margin analysis across business dimensions. Python Pandas SQL Power BI Excel Revenue · cost · margin KPIs Window functions Dashboard-ready measures
Business problem

Financial performance requires a view of both revenue generation and the costs required to generate it. This project examines performance across revenue, expenses, profitability, and margins to support management reporting and decision-making.

Dataset

Financial data covering revenue, expenses, and margin performance across business dimensions.

Analysis
  • Analysed revenue and expense patterns across business dimensions.
  • Evaluated profitability and margin performance alongside absolute profit.
  • Used SQL window functions to calculate financial KPIs and comparative period metrics.
  • Built reusable calculations for revenue, cost, profit, and margin analysis.
Key findings
  • Revenue growth is assessed together with cost performance, so growth is not confused with profitable growth.
  • Margin analysis identifies which dimensions contribute profit rather than only volume.
  • Financial KPIs are converted into dashboard-ready measures with a consistent calculation basis.
  • Analytical calculations are combined with visual reporting for management-oriented interpretation.
Business value

Gives management reporting one consistent view of revenue, cost, and margin together, so performance is judged on profitability rather than revenue alone.

View code & documentation on GitHub
Sales forecasting dashboard showing historical monthly sales and a forecast with error metrics

Sales Forecasting & Time Series Analysis

Retail sales forecasting with chronological validation — and an honest error rate attached to it. Python Pandas SQL statsmodels Matplotlib Excel Power BI 1,033,036 cleaned records ~12% monthly MAPE Chronological validation
Business problem

Retail businesses need reliable sales forecasts for planning inventory, staffing, purchasing, and revenue expectations. This project analyses more than one million retail transaction records and builds a time-series forecasting workflow from historical sales patterns.

Dataset

UCI Online Retail II — 1,033,036 cleaned transaction lines after preparation and exclusions.

Analysis
  • Cleaned and prepared large-scale retail transaction data, aggregating transactions into a chronological sales series.
  • Evaluated forecasting approaches using chronological (time-based) validation rather than random train-test splitting.
  • Measured forecast performance with monthly MAPE as the error metric.
  • Documented resulting KPIs and dashboard requirements for Power BI and Excel reporting.
Key findings
  • A seasonal-naive forecasting approach achieved approximately 12% monthly MAPE for the analysed forecasting setup.
  • Random train-test splitting gives an unrealistic view of forecast accuracy on time-series data; chronological validation is required.
  • Cleaning decisions matter at this scale: the workflow reduces the raw transactional data to 1,033,036 analysable lines before the series is built.
Business value

Gives planning teams a defensible baseline forecast with a measured error level attached, so inventory, staffing, and purchasing decisions are made against a stated expectation rather than an assumption.

View code & documentation on GitHub
Customer segmentation dashboard showing RFM segments and revenue concentration

Customer Segmentation

RFM segmentation on 1.07 million transaction lines — which customers actually carry the revenue. Python Pandas SQL Power BI 1.07M transaction lines Top 10% = 63.2% of revenue Gini 0.737 concentration
Business problem

A UK-based online giftware retailer serves a mix of consumers and wholesale buyers. Management can see total revenue but not which customers generate it. The analysis segments customers by value and behaviour to support targeted marketing and retention decisions.

Dataset

1.07 million transaction lines from a UK-based online giftware retailer, covering both consumer and wholesale buyers.

Analysis
  • Cleaned and prepared more than one million transaction lines for analytical use.
  • Defined commercial KPIs: customers, orders, revenue, average order value, and purchase frequency.
  • Applied Recency, Frequency, Monetary (RFM) scoring to classify customers into segments.
  • Used SQL to validate analytical calculations and Power BI to communicate the resulting segments.
Key findings
  • The top 10% of customers generate 63.2% of net revenue; the top 1% generate 31.1%.
  • 272 customers — 4.6% of the customer base — account for half of total revenue.
  • The revenue distribution has a Gini coefficient of 0.737, confirming substantial concentration.
  • Champions make up 21.9% of customers but generate 69.3% of revenue, making them the highest-priority segment for retention.
Business value

Shows where retention and marketing effort returns the most: a small group of customers carries the majority of revenue, and the segments to target first are named and measurable.

View code & documentation on GitHub

More projects

The same end-to-end workflow applied to HR, SQL, and operations.

HR analytics dashboard showing employee attrition, job satisfaction, and workforce KPIs

HR Analytics

Employee attrition analysis — which workforce factors are associated with people leaving. Python Pandas SQL Power BI Excel 1,470 employee records 16.12% attrition 65 SQL queries cross-validated
Business problem

HR management can see headcount and total turnover, but not which workforce factors are associated with employees leaving. This analysis investigates attrition and converts the findings into practical retention priorities.

Dataset

IBM HR Analytics Employee Attrition & Performance — a public dataset of 1,470 employee records covering satisfaction, training, performance, income, and tenure, analysed as a fictional case-study company.

Analysis
  • Data cleaning and consistency checks across 1,470 employee records.
  • Exploratory analysis of attrition against satisfaction, training, performance, and income.
  • 65 SQL queries cross-validated against Python and Pandas calculations.
  • KPI analysis and a 3-page Power BI report with interactive filters and synchronised slicers.
Key findings
  • Overall attrition is 16.12% — 237 employees out of 1,470 in the dataset.
  • Attrition is 22.8% for employees with a job-satisfaction score of 1, compared with 11.3% at a score of 4.
  • Attrition is 27.8% among employees with no training courses, compared with 9.2% among those who completed 6.
  • Correlation is kept separate from unsupported assumptions, so findings are presented as associations rather than proven causes.
Business value

Gives HR a prioritised view of where attrition risk concentrates — satisfaction and training exposure — so retention effort can be targeted instead of applied as a blanket policy.

View code & documentation on GitHub
SQL analysis output showing customer cohorts, retention, and revenue concentration queries

SQL Business Analysis

Advanced SQL on a relational business database — from data quality checks to cohort and LTV analysis. SQL Python Pandas Power BI 10 documented SQL scripts CTEs · window functions Cohorts · LTV · Pareto
Business problem

Business data often contains quality issues and requires more than simple aggregation to answer commercial questions. This project moves from raw relational data to validated business analysis using advanced SQL.

Dataset

Relational business database covering customers, orders, order items, products, and regions, analysed through 10 documented SQL scripts — from data quality through to advanced analytics.

Analysis
  • Audited data quality and validated analytical assumptions before reporting.
  • Multi-CTE queries with NTILE window functions to build RFM scores directly in SQL.
  • Sequential LAG/LEAD analysis, recursive CTE calendar series, and rolling 3-month moving averages.
  • Churn detection using subqueries and EXISTS, market-basket affinity pairs, and matrix pivoting.
  • Cross-checked SQL outputs against Python and Pandas calculations.
Key findings
  • Advanced SQL constructs are applied to real business questions: cohort retention, customer lifetime value, revenue concentration, and churn.
  • Cohort and Pareto analysis identify how much of the business depends on a small share of customers.
  • SQL results are cross-validated against Python and Pandas before they are used for reporting.
  • Results are translated into a Power BI reporting structure suitable for business stakeholders.
Business value

Demonstrates the database work employers test directly: analytical SQL that is readable, validated, and reusable — not just single-table aggregations.

SQL excerpt — sql/10_advanced_sql.sql
WITH customer_rfm_raw AS (
    SELECT
        c.customer_id,
        (CAST('2024-12-31' AS DATE) - MAX(CAST(o.order_date AS DATE))) AS recency_days,
        COUNT(DISTINCT o.order_id) AS frequency_orders,
        SUM(oi.line_total)         AS monetary_spend
    FROM customers c
    JOIN orders o       ON c.customer_id = o.customer_id
    JOIN order_items oi ON o.order_id    = oi.order_id
    WHERE o.order_status NOT IN ('Cancelled', 'Returned')
    GROUP BY c.customer_id
),
scored_rfm AS (
    SELECT
        customer_id,
        NTILE(4) OVER (ORDER BY recency_days DESC)    AS r_tile,
        NTILE(4) OVER (ORDER BY frequency_orders ASC) AS f_tile,
        NTILE(4) OVER (ORDER BY monetary_spend ASC)   AS m_tile
    FROM customer_rfm_raw
)
SELECT * FROM scored_rfm;
View code & documentation on GitHub
Supply chain dashboard showing inventory levels, demand, and logistics KPIs

Supply Chain & Inventory Analytics

Operational analytics for inventory, demand, logistics, and supply-chain performance. Python Pandas SQL Power BI Inventory · demand · logistics KPIs SQL-validated metrics Dashboard specification
Business problem

Supply-chain teams need visibility into inventory, demand, logistics, and operational performance. This project applies business analytics techniques to identify patterns that support better inventory planning and operational decisions.

Dataset

Operational supply-chain data covering inventory, demand, and logistics records.

Analysis
  • Examined inventory and demand patterns together to compare stock levels against demand movement.
  • Analysed operational and logistics performance as structured KPIs.
  • Used SQL to structure and validate supply-chain metrics before reporting.
  • Prepared analytical outputs as a Power BI dashboard specification.
Key findings
  • Inventory and demand are analysed together, so stock positions are reviewed against demand rather than in isolation.
  • Operational and logistics performance is converted into measurable KPIs that can be tracked over time.
  • Supply-chain metrics are structured and validated in SQL before they reach a reporting layer.
  • The output is a dashboard specification, not an ad-hoc chart set, so the analysis can be repeated on new data.
Business value

Converts operational data into a repeatable dashboard specification that supports demand planning, inventory control, and logistics review.

View code & documentation on GitHub

05 · Experience

Experience

Professional activities across data analysis, teaching, and independent work.

  1. Independent Data Analysis

    Independent · Oct 2021 – Present

    Self-directed data-analysis work applying Python (Pandas, NumPy, Matplotlib), SQL, Excel, and Power BI to explore datasets, build dashboards, and communicate findings.

    • Cleaning and preparing datasets for analysis
    • Building exploratory analyses and visualizations
    • Designing dashboards and reports in Power BI and Excel

  2. English Teacher

    Independent · Oct 2019 – Feb 2021

    Teaching English language with a focus on clear communication, structured materials, and measurable learner progress.

    • Planning and delivering lessons
    • Preparing learning materials and assessments
    • Communicating complex ideas clearly to diverse audiences

  3. Freelance Work

    Jul 2019 – Present

    Freelance and independent work combining data analysis, writing, and clear communication.

    • Analysing large datasets to uncover trends and solve business problems
    • Building dashboards and reports that make complex data easy to understand
    • Data cleaning and preparation to keep datasets accurate and analysis-ready
    • Writing SQL for database querying and data manipulation

06 · Education

Education

  1. Master's Degree

    English Language and Culture Abbas Laghrour University — Khenchela Sep 2019 – Jul 2021

  2. Bachelor's Degree

    English Language and Literature Abbas Laghrour University — Khenchela Sep 2016 – Jul 2019

07 · Certifications

Certifications

Courses, learning paths, and professional certificates from LinkedIn Learning, Microsoft, Udemy, and Astronomer.

  • Data Analytics

    Career Essentials in Data Analysis by Microsoft and LinkedIn Microsoft · Jun 2023 View credential

  • Business Analysis

    Career Essentials in Business Analysis by Microsoft and LinkedIn Microsoft · Dec 2024 View credential

  • Professional Certificate

    Data Engineering Foundations Professional Certificate by Astronomer Astronomer · Dec 2024 View credential

  • Python

    Mastering NumPy, Pandas and Matplotlib - Data Manipulation Tool Udemy · Jan 2026 View credential

  • SQL

    SQL for Data Analysis LinkedIn · Dec 2024 View credential

  • Excel

    Mastering Microsoft Excel: A Comprehensive Guide to Boost Your Skills Udemy · Feb 2025 View credential

  • Power BI

    Power BI Essential Training LinkedIn · Jul 2023 View credential

  • Tableau

    Tableau for Data Scientists LinkedIn · Dec 2024 View credential

  • Learning Path

    Become a Data Analyst LinkedIn · Jul 2023 View credential
All certifications 80 credentials

Data Analysis & Visualization

SQL & Databases

Python & Programming

Excel & Spreadsheets

Power BI & Tableau

Business Analysis & Project Management

Data Engineering

Statistics, AI & More

08 · Contact

Get in touch

I'm open to data analyst and business intelligence roles. The fastest way to reach me is email — I usually reply within a day.