We always envisaged Effective Machine Learning Teams (EMLT) speaking to multiple roles. Recently we’ve spoken to 4 distinct audiences about ML development, and you can listen in too.
Whether you identify with:
- Product
- Engineering
- Data
- Organisation
- more than one of the above
there’s a podcast* for you with popular presenters – see more below.
Product
We joined a Product Tank Melbourne event, along with Ana Kelk from Canva (and hosted by the wonderful Si Chen) to talk about how product discovery, development and management practices can be extended into the ML world, and the importance of taking a product mindset to building ML solutions.
* this is the one conversation we haven’t recorded – product podcast recommendations welcome!
Engineering
We talked with Henry Suryawirawan, generalist software engineer, on his excellent pod Tech Lead Journal. We explored ML issues that software engineers might face, and how good development practices extend to ML, including building AI products, and deep discussions about testing.
Data
We talked with data engineering legend and “recovering data scientist” Joe Reis on his podcast. In this instalment we did a deeper dive into data management practices that support ML, and especially the wastes that are all too common in data processes and data organisations.
Organisation
We talked with Matthew Skelton, co-author of the highly influential book Team Topologies, on Stream of Teams podcast. We talked about the technology and practices within teams and the team shapes and interaction modes between teams that support fast flow in ML development.
Keeping the conversation role-ing
This is by no means the end of the conversation, and we know these four perspectives don’t cover all roles, so we look forward to more conversations about effective teams in ML and AI.
* Ada was lucky enough to be enjoying an extended holiday during the period David and I recorded the Engineering, Data and Organisation podcasts
Leave a Reply