Engineering Management Q&A and Videos

Context:

(1) My team owns a service for which we're rolling out a new version with a big revamp of all the public interfaces and a ton of breaking changes.

(2) This is a legacy system that is being refactored to resolve some severe issues that its consumer systems have been complaining about for a long time.

(3) This service has many consumers in our org across multiple teams that depend on it for a lot of critical functionality.

(4) We need to migrate all consumers to the new service. My team cannot parallely support both the versions and the legacy system has to be deprecated before the new service deployment.

Challenges:

(1) Originally, the plan was for my team to roll out the new service and migrate all of the consumers to the new service as well.

(2) Now, we've had a huge scope expansion in the refactoring itself due to which the project timeline has extended massively.

(3) My team feels that working on such a long timeline project is risky and prone to further scope expansion if new consumers start using the old legacy system in the meanwhile.

(4) Another challenge with this is that my team has no context or understanding of all the consumer systems.

Questions:

(1) What approach can I use to now change the plan and convince the managers/tech leads of the consumer teams to own the migration of their consumers code to the new service?

(2) In general, what approach would be ideal for such a large-scale migration - Centralized migration by the service provider team vs distributed migration by all the respective consumer teams?

Hey Taro,

I have a question around how to manage a direct report who I’m receiving performance complaints about.

For context, I’m the team lead for a team that works distributed across vertical teams. We have a new hire who’s been working with us for around 3 months now. He’s working with Team A and has recently been helping them to write App X, a brand new application.

Recently the PM for Team A has reached out to me with some general complaints/concerns about his output. The work this new hire is doing blocks most work for Team A, and the PM feels like they should have finished their work by now, and apparently other members of Team A have raised similar complaints.

Additionally, this new hire has a reputation for working very long hours/late into the night after work. On several occasions he’s posted slack messages at 1am, and the PM is concerned that he’s trying to spend long hours after work to try and make up for his lack of progress.

Myself and my manager have both already had casual conversations with him about his late night work to try and put a stop to this, and myself and other team members have tried to help him with his tasks where we can. We’re in a very small company so while I can try shuffling him around to a different vertical, it’s not like there’s limitless possibilities there.

I’ll be bringing this up with my manager today during our 1:1, but I was also very interested in hearing what the Taro community thinks here.

This is my first time as a manager so I’m very much out of my depth here.

I had taken AI/ML over a decade ago in College.

Was wondering how folks have transitioned to the space. Particularly from management perspective.

A teammate of mine has sometimes delayed tasks and informed about them at the last minute leading to us missing the delivery deadlines. As the tech lead, I've shared this feedback along with explaining the impact and suggestions to tackle the issue. I've encountered this situation a few times even after sharing this feedback.

What other steps can I take to help the teammate and minimise the overall impact on our delivery?