My job is mostly to do with various technical studies involving statistical applications in network engineering, for example capacity planning. I frequently do other tasks for senior management where statistical expertise is required. I have done work for * CFO, Openreach, regarding the estimation of liabilities* Group Customer Experience, introducing statistical tools for looking at KPIs, predicting customers who will churn,* Group HR, data analysis in support of making sure that our people are treated fairly and equally, irrespective of their gender, age, ethnicity or employment status, * Group Regulatory Finance, looking at producing estimates of the amount of network assets for the BT Financial Reporting Statements, * Subjective Voice Quality engineers, help with analysis of designed experiments. * I am a tutor on the annual Data Science course. I usually like to teach R and Exploratory Data Analysis modules and anything to do with Bayesian statistics.

Researched highly automated telecare systems that looked after a frail house occupant living alone. Automatically detected behaviours that indicating a situation of concern and called for help if necessary without the need for active intervention by the occupant.

Developed an 'early warning' system for Major Accounts that automatically warned if calls or paid minutes used by major accounts increased or decreased unexpectedly. Researched tree models, both classical and Bayesian. Gave regular presentations to major accounts on the benefits and pitfalls of data mining.

Provided statistical support for BT's marketing databases. Analysed the effectveness of BT's 'Animals' and 'Maureen Lipman' advertising campaigns.

Developed practical applications of the PROGOL ILP language to the telecommunications business.

Taught the quantitative modules to Diploma in Management Studies (DMS) and Master of Business Administration (MBA) students. The courses included Use of spreadsheets, Optimisation, Inventory Control and Just-In-Time, Forecasting, Utility theory.

Optimised BT's basket of prices to maximise profits under the RPI-X economic regime imposed by the regulator.

Rewrote the Bayesian forecasting module of the MEMIS marketing database for run-time efficiently. The original version (written in BASIC) took over 10 hours to run. My Fortran 77 version took 10 min to run and was a lot easier to maintain because of its simpler data file structure. The code was re-used in several other projects that need forecasting.