Jump to content

Amena Zepherin

  • Posts

  • Joined

  • Last visited

  • Days Won


Everything posted by Amena Zepherin

  1. My name is Alex, an MSc Data Science postgraduate student with Robert Gordon University in Aberdeen and member of the teams that achieved 2nd and 3rd place in Project:Hack 8 and Project:Hack 9 respectively. Having graduated in law, worked in banking and asset management, and then pursued a master’s degree in data science, I love walking out of my little domain bubble. What are other companies doing, what are other industries doing and how are they doing it? You never know what you are going to learn or who you are going to meet, and information is valuable. It gives you perspective. These questions brought me to Project:Hack 8 in March of this year and, subsequently, Project:Hack 9 in June. Both were well organised and structured by Projecting Success, with technical support ready if needed. The challenges were clear, practical, and commercially relevant, and the pool of participants diverse. From students, apprentices and newly minted coders to data scientists, project managers and the wonderful people from Gleeds: James Garner, Nicola Herring, Manojit Sarkar, Nahid Jafar, Sheldon Atkinson and Basel Yousef, who welcomed and treated me like an equal member of the team although I was just a student. During Project:Hack 9 we tackled Challenge 11, namely a cost prediction model that uses machine learning to improve cost forecasting on construction projects. Whilst this challenge was within my field of data science, I had limited knowledge of the context in which the solution was being developed, namely construction projects, and two of the programming languages. But I was curious and ready to put in the effort. We were pragmatic and creative in our approach, a reflection of Gleeds’ own vision and values, and within 18 hours we created an innovative solution that brought us 3rd place. As well as working in HTML and Java for the first time, and more interaction with Power Apps, I also gained deeper insight into an industry that I had previously thought as difficult to access given my background but which I am now actively considering shifting to from financial services. As with everything in life, an opportunity is only as valuable as you make it. Whether looking to put your skills into practice, expand your portfolio or strengthen your business acumen for graduate programme applications, Project:Hack is a well worth investment as a student. Be brave, be intrigued and be yourself. The challenge sponsors want to work with you, they want to hear your ideas and see what you can create. Your input matters. As for me? I’m a convert. With two successful hacks behind me and many challenges still to be tackled my work here is not done. Prior commitments are keeping me busy in August however roll on Project:Hack 11!
  2. If you’d ask me 6 months ago what a data analyst did, I would've given you a shrug at best, but after the intensive week involving a crash course in Python Scripting and Power BI followed by my first Hackathon, I think I’ve turned into a data evangelist. Apprentices can take advantage of the Government Apprenticeship Levy, enabling them to get advanced analytics training at no cost to themselves or their employers. I previously spent two days at Project:Hack as a complete novice. Without a lot of tech skills to offer my small team, I focused on the User Stories, dashboards and the presentation. But what I saw being developed by the hundreds of people in attendance was staggering. As a more mature student with a well-established role at Costain, I don’t really see myself becoming a full-time data analyst. What I would like to do with the skills I gained from the apprenticeship is take the new skills back to the projects to produce and share useful project management tools. Also I see a side-quest in championing a Data-as-an-Asset movement. Data is much like any real physical asset. It can be optimized in vast quantities, learned from and focused onto real world outputs. To do this we need people to translate Client Needs into problem statements that have answers lurking inside of data. This is where I see myself operating. Andrews' foresight for data analytics is inspiring. Every intake at the academy demonstrates to us the value in bringing data analytics to project delivery and it's great to see them go on to pioneer change within their organizations.
  3. Hi Everyone, A huge thanks to everyone who supported and participated in Project:Hack9! I have attached the consolidated feedback and scores from all of our judges. We hope that you find the feedback helpful in preparing for Project:Hack10. Hope to see you at the next one! Dates to be announced soon. What do you think? What would you do to improve? Feedback Hack9.xlsx NOTE: To download the feedback, please join the Project:Hack club
  4. I’m loving learning about data analytics in a well-structured way. I’ve tried learning more about data analytics before but there’s so much information out there and it’s difficult to separate the good from the bad. Before the apprenticeship, I struggled with Data visualisation and Python but now I have developed a great love for both. The tutors explained all the concepts in an understandable way and the lessons were very rewarding. I have already completely changed careers to be much more involved with data analytics and I believe this is all down to the apprenticeship. It has given me the key skills to be an effective data analyst and it’s also given me the grounding to then be able to look at other learning opportunities and add them to my knowledge base. My challenge required analysing material requisition data from a large infrastructure project. I wanted to get an understanding of what materials prices, price variations and ultimately try to develop a tool which could identify and alert when the best time to buy was. A couple of data scientists joined the team and we seemed to be making good progress until we realised the data quality was bad. The project ended up being an exercise on data quality, how to improve it and what impact the poor data quality had on the final output.
  5. Challenge Overview Can we develop an accurate and robust machine learning model to improve cost forecasting on construction projects? We would then like to integrate this model with our existing dashboard. Challenge Breakdown A challenge that builds on the work of the Knowledge Transfer Partnership (KTP) with the University of Nottingham. This challenge intends to improve cost forecasting and/modelling to make it an automated and efficient way to reduce cost. The challenge is to create a machine learning prediction model and further develop our dashboard to allow automatic access to the predictions. The data itself is from the BCIS a paid subscription service. The data provided will just be for the hackathon itself. You will be developing a predictive model using project level information. Skills: Learn Power Bi through an existing dashboard Develop your programming skills (for machine learning specifically) Learn Power Automation and create work flows
  6. Challenge Overview Can we utilise GPT-3 to create tailored messages to different stakeholder groups based on contextual project information and message templates? Challenge Breakdown A challenge that involves using the power of a state-of-the-art natural language model (GPT-3) to automatically create tailored messages to different stakeholder groups on a construction project. You will need to create personalised messages to the various stakeholders using the data provided as prompts to GPT-3. You are required to 'engineer' the prompts that you provide the model to achieve the desired outputs. Skills: Working with APIs Using the GPT-3 natural language model Investigation into data-driven stakeholder management
  7. Challenge Overview Can we use the power of graph databases and graph analytics to identify weaknesses that exist within the supply chain of a portfolio of projects? Challenge Breakdown A challenge to find supply chain weakness. You will be using US government data to develop complex graphs using Neo4j or something similar. With focus on graphing, this challenge allows you to explore the more advance visualization options while still challenging you to keep it relevant and informative. You will need to identify critical metrics of the supply chains and reveal the most vital elements. Skills: Use Neo4j A focus on complex graphs Work with a large dataset
  8. Challenge Overview Can we use gitHub's API to search through existing repositories to determine how long a programming project will take to develop and how likely it is that there will be delays in development? Challenge Breakdown A challenge about predicting how long a programming project will take. This challenge requires you to gain meta data from GitHub a free programming repository. Use the huge data set to create a search function to be able to see how long certain types of projects take to develop. Use tags and whatever data you can find to inform these searches to provide as much insight as you can get. Skills: Web scrapping and API handling Learn Power BI Work with a massive and varied set of data Find your own insight and decide what is important
  9. Challenge Overview Traditional project plans regards dates and durations as being known and deterministic, a simplified model which often struggles when it encounters the realities of execution. This challenge is focussed around the opposite - obtaining insights from a plan where all dates and durations are described in terms of probability distributions. Challenge Breakdown A challenge to find insight into milestone data and nondeterministic durations. The data set for this challenge contains task information and how long it is predicted to take, compare this with a variable result and find insight on where bottlenecks and issues occur. Use decision topology information and create your own metrics to present the data in a meaningful way and perhaps find additional insight into what is important. Skills: Learn Power BI Use data to inform about project management Learn and develop new visualisations for project management Create your own metrics
  10. Challenge Overview Can we perform a detailed investigation into the design performance on our portfolio of projects, developing a set of metrics based on project financial performance. We would also like to understand in detail the performance on individual projects, to allow Project Managers to implement mitigating action. Challenge Breakdown A challenge to measure the performance of a project and where it can be optimized. In this challenge you will have to combine various data sets to establish a metric for project performance. We will provide some performance metric ideas but it will be up to you to get useful insight from the data. Take RFIs, drawings and project level infomation to create dax or python queries to find the insight you need. Skills: Handling and linking multiple data sets Learn Power BI and DAX Find new insight in the data Build meaningful visuals for project managers
  11. Challenge Overview: Can we create a tool to flag potential concerns within the cost budget for a project based upon forecast and actual cost data from previous projects? Challenge Breakdown: A challenge about using past data to see where cost forecasts may be inaccurate. This challenge will have you using the provided data to try and find trends in unexpected purchase costs. Then you will use this metric to warn staff if their future predictions do not match what has happened in the past. We want you to try and find deeper insight with this challenge and provide us with ways to take full advantage of this data. Skills: Finding insight in limited data Developing metrics to predict what may happen in the future Learn Power BI Explore new and relevant ways of helping businesses prepare for new projects
  12. Challenge Overview Can we create a reporting tool to receive weekly updates on the mental health of our teams throughout various projects? Can we gain insight from this data and create a flow that keeps management appraised of our teams' mood and needs? Challenge Breakdown A challenge to explore the topic of mental health in the working world. This challenge is about developing a mental health reporting system for various companies to use to track their staffs mental health in relation to project management. The task is open ended and we are very excited to see what may be possible in this space. Skills: Learn Power BI, Power Apps and Power Automate Develop a solution that will act as a foundation for the future of mental health reporting Work on new ideas to explore new possibilities Develop the full flow of a new solution
  13. Challenge Overview Can we efficiently display relevant visuals for all our health, safety and environment data that is easy to understand and tailored to employees at different roles? Challenge Breakdown A challenge to develop four unique dashboards in an attempt to set an industry standard. Work with various different data sets to try and tell a story and create reusable dashboards. This challenge is all about creating the insight. The four dashboards are about HSE metrics and developing a method to provide all the health and safety data to a project manager and higher management. Using data about incident reports, waste and carbon emissions help to improve safety in multiple industries. Skills: Learn Power BI Create vital insight for multiple industries Develop a solution meant for project managers and learn to tell a story with your data
  14. Challenge Overview Can we scrape project cost estimates from an unstructured data source into a structured form? We will be developing a visualised search tool to be able to access this data in a quick and efficient manner. Challenge Breakdown Budgeting and estimation in the construction industry is a vital but often a time-consuming endeavour. Can we build a testable dashboard to search through this data and gain further insight into the data through visualisation? This challenge is about making it easier to work with unstructured data to search and find information. The challenge will require you to take a pdf filled with data and tables and pull that all into a completely new structure. You will then have to provide a search feature to allow industry professionals from a non-technical background, the opportunity to search for the things that they need. In this challenge you will develop your critical skills through collecting data from unstructured sources, developing queries and learning the ins and outs of the Power BI and Azure Cognitive search features.
  15. We all know projects are extremely complex system that are influenced by extremely large number of factors. Often project failure can be due to hidden causes that can sneak up on project managers. Data holds the key to preventing this, as it allows us to peer into the past and better understand our mistakes. In many projects, one of the main major issues holding us back is that we lack the quantity of useful data needed to perform in depth project data analytics and the data we do have is limited in scope. Currently, project managers mainly pull from 1st party datasets (Datasets with a direct relationship with the problem), which does provide some insight. However, do many people know of the true power of 3rd party data? What is Third Party Data? Third party data is data that sits outside the organisation. Often the data has no direct relationship with problem at hand. A good example would be when a project manager wants to know why there are delays in the a construction project, they may use weather data to provide deeper insight into the condition on site. In doing this project managers can have a holistic view of the project. Project professionals may find this is interesting, however are there any real world examples of third party data being used? Well the winners of Project:Hack 8 (Challenge 5a) displayed the power of third party by solving this problem. Here cost managers need to understand price trends in key raw construction materials, in order to know when is best to buy them. So the team took data on construction material from the Office of National Statistics and third party data from the London stock exchange, to provide insight into the price of these raw materials. This allows cost managers to make more informed decision, when buying these materials and also it allows data scientists to better machine learning models that make more accurate predictions. AWS Data Exchange In the previous example, code had to be developed and deployed to allow for the extraction and integration of the third party data. This process is time consuming and costly. AWS Data Exchange saves on these cost greatly, as this service provides pre-built data pipelines connecting data subscribers (consumers of data) and providers (producers of data). The service acts as a place where data subscribers can gain easy access to third party data. Meanwhile, for data providers it acts as a market place, where they can sell their data on. This enable through the AWS Data Exchange API. Project professionals and project data analysts will have access to data from all across the world and across industries. The marketplace offers useful data sets including data on COVID-19 to data from the public sector. These new data source can help expand the scope of project data analytics and thereby allow management to make more informed decisions. How does this all work? Once sourced, this data can be then stored in Amazon Simple Storage Service (Amazon S3), which is a scalable storage service. This data can be further analysed with AWS analytics and Machine Learning services to provide important insights. So when looking to advance your analytics, think about third party data sets and how they can enrich your data.
  16. In our most recent Meetup, we had a talk from Highways England on how they are transforming their approach to project delivery. With deployment of advanced data analytics across various stages of their projects including the A14 and Lower Thames Crossing, data analytics is becoming a standard in how they deliver projects. Another step into their data journey, includes upskilling their existing project workforce to align with this new approach, of which they have utilised the Project Data Academy to achieve this. Hi, and welcome to my blog. By way of introduction. I am Vicki Perkins and I work for Highways England, in the Complex Infrastructure Programme as a Digital Project Manager. My main role is working on the Data Analytics and reporting project we are currently running. This takes the lessons we learnt from A14 Cambridge to Huntingdon of using an analytics reporting platform to gain insights and decision making in a timely manner and applying these to the other 3 projects (Lower Thames Crossing, A303 Stonehenge, A428 Blackcat to Caxton Gibbet). How I Am Adapting To A Project Data Analytics Future As A Project Professional Having followed The Project Data Analytics Community for a few months now, its clear that the future of project delivery will be vastly driven by advanced data analytics. Not only because data is inherently better at informing our decisions, but there seems to be a real buzz around concepts such as automation, low code no code, artificial intelligence and machine learning. Being a project professional with some data knowledge. I know advanced data analytics is a broad concept that I will never be able to understand in its entirety. But to help optimise my role, I wanted to be able to find a niche within data analytics and understand it to a level where I can implement long-term solutions to our projects. This in turn will enable me to help drive the data capability within my organisation and prepare us for a data-driven future. As opposed to self-taught learning, I joined the Project Data Academy run by Projecting Success, where I am working towards a Data Analyst Lv4 Apprenticeship. I am a recent start so my thoughts on the matter reflect what its like to cast a net into a wide pool of data, with hopes that it will train a foundation for advanced project data analytics, but also guide my interests into this vast phenomena. My first experience at the Project Data Academy, was at the intensive week. Before entering into this session. I had mixed feeling as to what to expect. I had been given an overview of the course but had little other information. I remember thinking I wonder how much of this is going to be similar to my day job? From this first week I could see there was some overlaps and also some parts that were going to be completely new to me. This made me feel more at ease as we went through the week. I also could see it was going to be a steep learning curve on some of the technical parts even for someone like me; with previous data experience. Conceptually I understood what was being mentioned but I wasn't sure of technically it worked. I was also very aware there were others on the course that would have very different experiences to me. The personal intro's everyone did helped to breakdown some of the barriers, its good to meet others from the industry and find out what their problems are and collectively how can we do things together. As the week progressed, it became clear that everyone was on a very different starting level. This would either encourage people along collectively or cause the quieter members to stay quiet for fear of saying something wrong or not understanding. After the intensive week, I went back to a mix of feelings from content that I knew what was expected of me and what was involved. but also worry and nerves as to 'could I actually complete this course'? A. from doing this alongside my normal busy day job, but B. do I have the capability to do the heavy technical coding etc bit. I needed to take some time out and reflect on this feeling. After a few days of careful thinking I had come round to the feeling of ‘yes’ I could do it and what an experience this could be. I learnt from the above experience to stop doubting myself and my abilities. Yes at times I am sure things are going to be tough, but I will find a way through it. I must not forget there are people to help and support me along this journey both from Highways England perspective to the team at Projecting Success. Through what I have learnt I can definitely apply this to my normal day job. This is very similar to the feelings I had following the first episode of learning. I need to believe in myself more and stop trying to doubt myself or my capabilities in data analytics.
  • Create New...