We specialise in data analysis and system development using SAS and SQL. Our service is reliable and efficient. Recent clients have been in the banking, marketing and retail sectors.
We design and develop:
- data processing (ETL) systems.
- processes and systems for statistical data analysis and reporting.
- exploratory data mining and ad hoc statistical investigations.
Typically our solutions involve: SQL code run against local or server databases; SAS code on local and/or server systems. The combination chosen depending on the client's installation. The solution would typically involve:
identifying the various data sources;
combining and merging datasets;
identifying errors and outliers, inputing replacement values if necessary;
cross-referencing to check logical integrity;
and then adding value e.g. through calculation of new metrics, summarisation, simulation and statistical tests
Our expertise covers:
- SAS Programming - developing datamart ETL systems using Base, Macro, Proc SQL, SAS/Connect, SAS/Access with back end connectivity to other databases
- Database Programming - e.g. Oracle SQL & PL/SQL, Teradata, SQL Server & T-SQL, DB2
- Statistical Analysis - SAS/Stat and SAS/ETL for linear regression, time series analysis and forecasting, analysis of variance and general linear models, logistic regression, cluster analysis.
- Reporting - SAS Graph, integrating SAS with Excel
- Integrating disparate data sources - PC Excel/Access/ODBC, Unix, Mainframe.
- Custom Algorithms / High performance code using C++ - e.g. optimisation, heuristics
Learn more about these in the technology overview.
Example subject areas:
- marketing: mailing campaign design and response analysis, propensity to attrite, lifetime customer value, price elasticity,survival analysis.
- retail: account profiling, customer base segmentation, rectification.
- finance & insurance: risk scoring, expected loss, stochastic simulations.
- supply chain: sorting and packing algorithms, network optimisation, heuristic techniques such as genetic algorithms amd simulated annealing.
Some previous clients
