Project - Exchange Rate Forecasting and Competitor Pricing Analysis

The Iranian branch needed to minimize currency conversion losses by predicting the optimal day to convert Dirham to Rial each week. Samsung required up-to-date pricing data for mobile phones in the Iranian market to track competitor strategies and maintain a competitive edge.

Client
Samsung - Iranian market research
Year
Service
Data Extraction and Scraping, Data Warehousing, Machine Learning and Forecasting

Overview

Exchange Rate Forecasting

  • Challenge: The Iranian branch needed to minimize currency conversion losses by predicting the optimal day to convert Dirham to Rial each week.
  • Solution:
    • Developed a forecasting model to analyze exchange rate trends and identify the best conversion days.
    • Managed the entire stack, including data analysis, model development, and web visualization, providing stakeholders with actionable insights.

Competitor Pricing Analysis

  • Challenge: Samsung needed up-to-date pricing data for mobile phones in the Iranian market to track competitor strategies and maintain a competitive edge.
  • Solution:
    • Built a data scraping system to collect pricing information from multiple sources and online retail stores.
    • Cleaned, aggregated, and stored data in a centralized warehouse for analysis.
    • Designed an interactive dashboard to present pricing trends, competitor fluctuations, and actionable insights.

Responsibilities

  • End-to-end data pipeline development, including:
    • Data extraction and cleaning.
    • Data warehousing and aggregation.
    • Interactive dashboard creation for visualizing market and currency trends.

Outcome

  • The exchange rate forecasting model enabled significant cost savings by optimizing weekly conversions.
  • The competitor pricing system provided valuable insights, empowering Samsung to make data-driven decisions in the Iranian mobile market.

What we use

  • Scrapy, Selenium
  • dbt
  • MlFlow
  • XGBoost, Prophet
  • Data Visualization and Dashboarding(Grafana)

More case studies

Employing LLMs for Incident Response in Telecom

Our latest project integrates Large Language Models (LLMs) with Knowledge Graphs to revolutionize Root Cause Analysis (RCA), improving the automation of incident investigation and resolution.

Read more

Predicting Stock Price Movements

Develop a robust predictive model to forecast stock price movements, supporting the sales department in maintaining and enhancing stock value.

Read more

Tell us about your project

Our office

  • UK
    Sanford House,81 Skipper Way,
    St Neots