AI for Business MSc postgraduate personal statement example

Personal statement example

At the hotel where I work on the reservations desk, every Friday afternoon someone prints a spreadsheet of the weekend's bookings and circles the names most likely not to turn up. The circles are drawn from memory and instinct. Sometimes they are right and we fill a room twice over with confidence; sometimes a family arrives at eleven at night to find no room, and I spend the next hour ringing other hotels. Watching this weekly ritual made me curious about how much of a business's judgement is already a prediction, and what happens when that prediction is made explicit, tested and explained to the people who rely on it. That question is why I am applying for postgraduate study in AI for business.

My degree in Business and Information Systems gave me a grounding in databases, process modelling and management accounting, alongside introductory programming in Python. For my final-year dissertation I worked with a small regional water supplier that allowed me to use an anonymised set of around four thousand customer complaint emails. I built a classifier to sort them into categories such as billing, leaks and meter queries, starting with a bag-of-words model and logistic regression before trying a slightly richer approach using TF-IDF weighting. The more interesting results were not in the accuracy figures but in the errors. Emails mentioning both a high bill and a suspected leak were routinely misfiled, and the staff I interviewed told me those were exactly the cases they most needed to see quickly. I finished the project convinced that a model's usefulness depends on understanding the decisions it feeds into, not only on its overall score.

Since graduating I have worked in hotel reservations for eighteen months. The job is ordinary, but it has taught me a great deal about how data is actually produced. I have seen how a booking channel records a cancellation differently from our own system, how staff override rates for regular guests without noting why, and how a messy field can quietly distort a monthly report. When our manager wanted to know whether midweek discounts were working, I pulled two years of booking exports into a spreadsheet, cleaned the duplicated records and produced a simple comparison by season. It did not settle the question, but it changed the conversation from opinion to evidence, and I was asked to repeat it each quarter.

Outside work I help my mother with the bookkeeping for her small dog-grooming business, reconciling payments and chasing invoices. It is unglamorous, but it keeps reminding me that most firms do not have data teams; any tool they adopt has to be cheap, understandable and trustworthy. I also play club chess, and reading about how engines changed the game, and how players learned to use them as training partners rather than oracles, has shaped how I think about people working alongside automated systems.

My reading has tried to balance technique and consequence. Andrew Ng's material on machine learning helped me understand gradient descent and overfitting properly, while Cathy O'Neil's Weapons of Math Destruction made me take seriously how opaque models can entrench unfairness when no one is able to question them. In a hotel setting, a no-show model that quietly penalised guests from particular postcodes would be both unfair and bad for business, and I would like to learn how to detect and prevent that.

I want a masters that strengthens my technical skills in machine learning and data handling while keeping them tied to strategy, governance and the realities of organisational change. I am a careful, practical worker who is comfortable with imperfect data and with explaining findings to colleagues who are not specialists. I hope to move into an analytics role in hospitality or services, helping organisations replace circles on a printout with methods they can understand, check and improve.