EVENT - Tuesday
Location120 Holborn, London EC1N 2TD, UK
When6:00 PM - 9:00 PM
DSF Startup Showcase with Secret Escapes
Join Data Science Festival – London in partnership with Secret Escapes. June 11th, we will be featuring 6 new and upcoming companies at our Start-up Showcase. Come and hear how these new companies use DS to solve real work problems, the issues their teams have encountered and also the mistakes and success that you should look for when you are starting your own projects.
Due to the popularity of Data Science Festival events, we are now allocating event tickets via a random ballot. Registering here enters you into the ticket ballot for the Data Science Festival Event at Secret Escapes on June 11th 2019, the ballot will be drawn on the 4th June 2019. Those randomly selected will then be e-mailed a Universe ticket for the event, with the joining details.
If you get an allocated Universe ticket, please bring a copy of your paper ticket or your ticket on your phone to the event to check in with your QR code. Tickets are non-transferable.
Please click here to apply for a ticket: GET TICKETS
6.00pm doors open
6.30-7:15pm talk – 3 Short Sharp Lightning Talks – Secret Escapes – Depop – Wefarm
7:15-7:45 pm – Refreshments
7:45-8:30pm talk – 3 Short Sharp Lightning Talks – Benevolent AI – OneSub – WinningMinds.ai
8:30-9.00pm – Close
Talk 1: Ross Gray – Secret Escapes challenges scaling Airflow to running hundreds of dynamically generated DAGs.
Bio: Ross started with Secret Escapes late last year. He has several years of experience as a software engineer working with Python, building a variety of things from websites and APIs to high-performance ticketing systems. Being new to the world of Data Engineering, he is enjoying learning the tools of the trade and the challenges involved in building out a first-class data infrastructure that can scale to meet the demands of a successful business.
Summary: The jobs in our data pipeline are either self-describing or built dynamically off of config files. As Airflow DAGs are simple Python objects we thought an elegant solution that’s loosely coupled with Airflow would be to generate DAGs dynamically based on our job config files/metadata. We’ve now got a couple hundred jobs with this implementation and we’re seeing performance issues where the scheduler is slow to assign new work. We anticipate hundreds if not thousands of jobs running in production when we reach maturity so tackling the high-risk questions of “Can Airflow scale?” and “Can Airflow scale micro batch pipelines with jobs running every X mins?” is a priority for us to answer. We’re therefore taking an experiment driven approach to tackling the question. We’d like to share that journey and our findings with the community.
Talk 2: Clemence J Burnichon – Depop QuickSearch recommendations
Bio: I joined Depop a year and a half ago as a senior data scientist and I am now leading the machine learning effort across Depop. Our mission is to generate knowledge and capabilities to improve our buyer’s and seller’s experience with Depop. During my 6 years as a data scientist, I have acquired experience in multiple ML fields such as computer vision, recommendation engine and NLP. Prior to Depop, I worked Net-a-porter and Sainsbury’s.
Summary: In this talk, I will present our latest feature: Depop QuickSearch that enabled us to personalise the in-app search experience for our community using graph processing and NLP.
Talk 3: Rob Stanley – Data Science in Languages I Don’t Understand
Bio: Rob is Head of Data at Wefarm, leading a team using machine-learning and data analytics to help connect every farmer to the advice, products, and services they need to grow, no matter their language.
Summary: 1.5 million small-holder farmers throughout East Africa (Kenya, Uganda, and Tanzania) have used Wefarm to connect with one another to solve problems, share ideas, and spread innovation — all for free, and all via SMS. At the heart of this platform is Machine Learning: helping us to efficiently route messages, and enabling us to recognise the intent and content of every incoming message including in languages not supported by any standard library: Luganda, Runyakitara (in addition to English and Swahili).
Talk 4: Jinwoo Leem – Harnessing data to revolutionise drug discovery
Bio: Jin is a bioinformatician at BAI, currently working on approaches to ingest and model data for enabling drug discovery activities. Previously he was at Oxford University for a PhD and postdoc, developing bioinformatics approaches for computationally designing antibody-based drugs.
Summary: Ten thousand papers are published in the literature every day, and biomedical data is accumulating at an unprecedented rate. Despite this influx of data, harnessing it to advance drug discovery remains an incredibly difficult problem. At BenevolentAI, we have created an end-to-end platform, starting from the foundations of data ingestion, all the way to drug design. In particular, our use of artificial intelligence ensures that the right drug is delivered to the right patient. It is our goal to disrupt drug discovery because it matters.
Talk 5: Jim Morrison – Breaking the Echo Chamber – AI versus Human Stupid.
Bio: Jim is founder of Deep Blue Sky, a small SaaS firm in Bath focused on bringing transparency, accountability and efficiency to our government’s International Aid spending. He is also director of the Bath Digital Festival, a charitable festival of digital tech bringing together 2,500 folk in the South West each October.
Summary: OneSub is a startup on a mission to bring balance and reason to news media. The echo chamber we’ve built for ourselves is destroying society and democracy. Can technology be used to save us from ourselves?
Talk 6: Dimi Masaouti – People Data for Good
Bio: Dimi (Dimitra) Masaouti is the Co-Founder of WinningMinds.ai, a conversational analysis AI platform that helps businesses build a team-centric productivity network. Previously, Dimi led people change and development programmes in different organisations such as Vodafone, Sanofi, GSK, RBS and others. In her past roles, she developed and productised solutions enhancing strategic alignment of teams to business imperatives as well as people performance. She holds a MSc in Strategic HR and an MIT certificate in Big Data and Social Analytics. Dimi’s peer, Maria, describes her as loud and non-stop, whilst Philippe, being gentle, said that she likes to push the limits.
Summary: Since the 1990’s workplace skills and behaviours are assessed with online tools and tests, which report on the user’s perception about his or someone else’s performance at a given time. In reality people’s behaviours and attitudes are time and context specific, they are not static. The assumption of stable personality traits and universal emotions is flawed and the theory of personality states and behavioural episodes emerge as more relevant and effective. Gathering and analysing real-time and context-specific data from business meetings generates more realistic and dynamic insights. WinningMinds mission is to enable people and teams develop their capability in action, analysing team meeting discussions and generating credible insights over time. To meet our promise we have taken a lot of actions to minimise bias and ensure that we generate reliable data that is used for good.