JP Morgan
Contact: "Elliott, Michael" <Michael.Elliott@jpmorgan.com> and "Agozino, Biko JJ" <biko.jj.agozino@jpmorgan.com>
Ideas for 2022
will be coming
Ideas for 2019
"Agozino, Biko JJ" <biko.jj.agozino@jpmorgan.com>
We certainly do have ideas from earlier in the year that we put aside for the Cambridge group projects. Let me see where they got to in the end and get back to you.
Ideas for 2018
Background:
The Royal College of Anaesthetists ran their own internal hackathon and came up with a few ideas. The idea they’d most like to take forward is a Perioperative Medication Advice app.
Currently, individual consultants / trusts have use their own experience guidelines around medicines (pre through to post) operation. This leads to incomplete and non-optimum information being provided to patients. For example patients taking anti-coagulants may be given differing advice as to when they should stop taking them prior to operation, possibly increasing the dangers through bleeding or result in more costly (to the NHS) blood transfusions. Often there is other physiological information that needs to be taken into account as well as just the information provided by the drugs companies. The number of possible drugs makes giving an individual patient specific, personalised perioperative drug advice difficult.
The desired solution would be reduce the reliance on human knowledge, leveraging the best practice, which currently reside in individual trusts’ policy documents. Specific, tailored information could then be given to individual patients.
My thought was we could run this as a challenge at CFG (our external hackathon) London, and as a Cambridge Undergad project next term. This could then lead into a FFG (our internal pro-bono effort) project next year, where we can make sure the development is heavily test driven / rigorous. The challenge is likely to be in getting sufficient info to build a prototype. When I bounced this off Biko, he suggested we could use Machine Learning to scrape info from any documentation (PDFs) RCoA can provide.
---
From my perspective the key challenges of this project will be:
1. Leveraging Machine Learning and Natural Language Processing techniques to scrape and rationalize sufficient data from PDF documentation detailing the guidelines
2. Creating an Application that will help patients and doctors get advice on Perioperative Medication specific to the patient I think this project could provide a Techincal Challenge, and Business Opportunity, as well as the obvious Human Interest.
What do you think? Could we mold this into something acceptable for the group projects?
Kind regards, Biko
2017 suggestion
"Many charities rely on volunteer networks, but small local charities sometimes don't have the staff to maintain those networks. On the flip side, many people would like to volunteer to help a local charity but aren't able to commit a set number of hours per week. Micro-volunteering allows individuals to offer up their skills on an ad-hoc basis and for charities to take advantage of that, and is extremely valuable to smaller charities or community organisations.
Your task is to design a micro-volunteering exchange, that can match volunteers to opportunities in their local area. Locality is critical in micro volunteering, perhaps a graph database (like neo4j) would be interesting here.
2016 project
earlier version
Michael says: "I don't know that we have street by street data, but let me find out."
Different areas of the UK, or of cities like Cambridge, have wide-ranging differences in average age. Your task is to help people understand these differences by changing the appearance of an online map and/or street view, to simulate the issues faced by the people who live there. You can start with an interface allowing the user to drive through Open Streetmap or Google Maps, but modify local parts to reflect the age of people who live in that area. For example, the appearance of text and graphics might be modified to simulate visual capability of older users, navigation controls might simulate fine motor control issues (shaking hands, etc), and street scenes modified to show effects of reduced sensitivity to light in night driving. The resulting interactive system can be used both to better understand age distribution, and to help younger users appreciate the needs of the older population in particular areas.
even earlier
The first idea would be an application or game with the theme of "How does it feel to be old?". Some initial thoughts from AgeUk on that are:
"For a lot of Gen Y (or even Gen X) who use the internet, apps, games, social media on smartphones or tablets all the time, every day, it is really hard to imagine how using this sort of thing might be a really different experience for an older person. Older people may have fine motor control issues (shaking hands, etc), eyesight and hearing issues, and even just issues in processing information at a slower speed. All of these things may affect how comfortable someone feels about using websites, apps, or even doing simple things on a smartphone. How can we create a tool or an app or a game that brings this to life for younger people – and maybe helps them to think about getting involved with Age UK or donating as an outcome? "
The second idea would be data analysis and visualization around data representing the problems faced by older people in different regions of the UK (e.g. those in rural areas may be more affected by issues of isolation, whereas those in cities by issues of crime and security). Again, a very early description
"How can Age UK represent our complex regional-based data about older people in a commanding, visual way? We have complex data sets (we can send) in spreadsheet form that map occurrences of many different issues for older people, e.g. regional stats on disease, mortality, loneliness, health, etc etc. Could we find a way to visually represent this in a way that might be useful for PR purposes, to show to potential funders or donors, in a way that brings to life the multiple issues that older people use, and the way that Age UK helps in all its 165 localities"
If you think these are promising, I will work to flesh them out into more fully fledged projects. I think the first has some interesting aspects of user experience design, and has good scope for delivering a fully featured end product. The second perhaps offers scope for some creative thinking about data representation.
