WeCrashed by gov.uk and MLInfra dive
Kate Reznykova
September 29, 2022
local news 📰
🇬🇧 Is Britain now in a full-blown economic crisis?
There is a prediction that the UK to enter a recession as early as this year. This is largely due to surges in inflation as the cost of living crisis impacts all demographic groups. However, the shape of any recession is more important to businesses and policymakers than whether a recession is recorded in the national accounts.
New mortgage rates now quoted at 5%-6%
HSBC sees fall in house prices into next year
tech-trending 📈
What is machine learning infrastructure, and why is it important?
The majority of companies are left out as they lack the knowledge and resources to build an efficient and scalable machine learning workflow and pipeline. The gap between players big and small is widening.
Recording data, tracking changes and managing different versions (a.k.a. "version control") are important aspects of keeping an organized and efficient workflow. Version control is becoming increasingly important as data scientists need to be able to reproduce experiments.
An automated machine learning pipeline might look something like this:
1. Data is collected from various sources and stored in a data lake.
2. Data is cleaned and prepared for machine learning.
3. Data is split into training and test sets.
4. A machine learning model is trained on the training set.
5. The model is evaluated on the test set.
6. The model is deployed to a production environment.
7. The model is monitored in production.
8. The model is retrained on new data as it becomes available.
The idea behind automated machine learning pipelines is to automate as much of the work as possible. This can make the process of building and deploying machine learning models much faster and easier.
Machine orchestration is a critical element of any machine learning infrastructure. It ensures that all of the various components of the system work together efficiently and effectively. The most advanced engines are able to optimize resource usage, predict future needs, and handle queuing effectively.
Inference deployment is the final piece of your ML pipeline deployment. You might not want to make it possible to deploy in a live production environment, but make sure your data science teams are self-sufficient when it comes to deploying new models for the software teams that will be integrating the predictive models into your business apps.
For example, a tool like Kubernetes can manage the orchestration of your application deployments and their dependencies, while a tool like Helm can simplify installing and managing Kubernetes applications. Suppose you’re looking for a complete solution for managing your ML pipeline deployment. In that case, you can check out the MLflow project, which provides a platform for managing the entire ML lifecycle, from training to deployment.
cash flow 🤑
Manchester-based Lunio nabs €15 million to enhance ad campaign security
London-based Armalytix has raised a further $1m (£920,000) for its anti-money laundering (AML) software that checks where funds originate from
Toqio has secured €20m (£17.84m) in funding, consisting of an €18.7m (£16.6m) Series A led by AlbionVC and a €1.3m (£1.1m) grant for its white label fintech platform
Doccla raises £15m to monitor patients remotely in ‘virtual’ wards
twitter spotlight 🔦
Still probably the best diagram about organizational structures, drawn ~10 years ago (and still relevant!) by @lmanul. Original source: bonkersworld.net/organizational…
— Gergely Orosz (@GergelyOrosz)
Sep 27, 2022
hot startup jobs🔥
tekion - senior-product-designer
sourceful - full-stack-software-engineer
pleo - performance-engineer
monese - senior-software-engineer
primer - senior-product-designer