From Authority to Leadership

Dr Kajee's leadership session encouraged me to reconsider what leadership looks like in practice. 

The first insight was that leadership is about taking responsibility for the collective good and using one’s influence to support others. This idea is particularly relevant to research. Academic environments contain many informal rules that may not be immediately visible to students or early-career researchers. Knowing whom to approach, how to ask for support, where to find an opportunity or how to navigate an unfamiliar institution can significantly affect a person’s progress. An inclusive leader does not treat this knowledge as personal capital to be protected. Instead, they make the system more understandable and accessible.

The session also made an important distinction between representation and inclusion. A space may contain people from different backgrounds without those individuals feeling that they belong or that their contributions are valued. This led me to reflect on the relationship between speaking up and leadership. Participants discussed moments in which they had witnessed uncomfortable behaviour but had struggled to respond because of hierarchy, social pressure or fear of making the situation more awkward. I recognised the tension between knowing what one believes and acting on those beliefs in real time. Good intentions are not enough if they do not translate into action.

As a Laidlaw Scholar, this challenges me to become more conscious of moments in which silence may reinforce an unfair dynamic. Ethical leadership does not require having the perfect response immediately, but it does require a willingness to question, intervene and learn. Courage in leadership may sometimes involve openly challenging a decision. At other times, it may mean checking in with someone privately, asking a difficult question or creating space for a person who has been overlooked.

Another significant theme was authenticity. Authentic leadership was described as the alignment between a person’s values and behaviour. It means acting consistently with one’s principles instead of performing a version of leadership that is based on popularity, imitation or external expectations.

This prompted me to consider my own core values and the extent to which they are visible in my actions. I value curiosity, fairness, empathy and integrity. However, these values only become meaningful when they influence how I conduct research, collaborate with others and respond to disagreement. Fairness may require acknowledging whose perspectives are absent from a research question. Empathy may involve listening carefully to the experiences of participants and colleagues. Integrity may mean being transparent about uncertainty, limitations or mistakes.

The discussion on feedback was also valuable. Leadership development depends on understanding how others experience our actions, not only how we intended those actions to be understood. This requires actively seeking critical feedback and resisting the instinct to become defensive. Not every piece of feedback will be accurate, but each piece provides information that can be examined. I want to become more willing to ask not only whether a task was completed successfully, but also how my approach affected the people involved.

This session helped me see leadership as a continuous practice rather than a fixed identity. It is messy, contextual and constantly changing. It requires moving between confidence and reflection, action and listening, independence and collective responsibility.

Moving forward, I want to focus on three areas. First, I will become more intentional about sharing information and opportunities with others. Second, I will pay closer attention to whose voices are heard or ignored within group settings. Most importantly, I will regularly ask for feedback from the people around me. 

The question I will carry into the next stage of my Laidlaw journey is: What are my core values, and how do I demonstrate them in practice?

From APIs to Database

This week, starting from a raw spreadsheet of around 2,200 organisations with most fields empty, my goal was to automatically populate key data points — websites, financial figures, employee counts, founding years, and sector classifications — using Python, public APIs, and large language models.

I began by auditing the data, writing analysis scripts to measure missingness across all columns, broken down by sector and postcode district. From there, I built a series of website-finding scripts. My first approach used the OpenAI API with web search, but it was expensive and only filled one column. I tried Google's Custom Search API next, only to discover that Google had deprecated open-web searching, making it unusable. After pivoting to Bing and fixing a silent bug that had been overwriting previously found results on every re-run, I realised the real opportunity: since the cost of a web search call is essentially fixed per company, I could ask for all ten empty fields at once instead of one. This gave roughly seven times more value per API call.

I then split the pipeline into two phases — a free Charity Commission API pass first, followed by an LLM pass only for fields that couldn't be sourced from official records. Both scripts colour-code cells by data source so provenance is visible at a glance.

Finally, I automated the consolidation of thirteen source tabs into the Master sheet, using fuzzy name-matching to detect near-duplicates, and built geographic visualisations including an interactive Leaflet map and a publication-quality population density choropleth.

The biggest lesson was that real data work is mostly about handling what you didn't expect — every version of the pipeline existed because the previous one hit something unforeseen.