Are you a hands-on engineering manager, and how do you make decisions about delegation?

I received the strong dissatisfaction from our business parter expressing strong dissatisfaction with our system's performance of anomaly detection system.  They were receiving 4-5 high severity ticket monthly out of which 60% being false positive.   I took three key steps to resolve this situation

1. Active listening and Investigation

- I immediately scheduled a meeting with the concerned stakeholders to understand their frustration.

- Worked with my team to conduct a thorough root cause analysis.

- Identify 3 primary technical issues.

a.  Aggressive threshold prioritising speed over accuracy

b.  System not accounting for planned processor maintenance windows

c.  Low transaction volume leading to statistical anomalies

2. Stakeholder Engagement

- Had a 1:1 discussion with the client leadership team

Discovered their core pain points.

- High severity ticket disrupting their team's work-life balance

- Total ticket count negatively impacting their team's performance metrics

3. Solution development and Implementation :  I proposed a balanced solution that address both team's need

Technical improvements

- Incorporated planned outage signals into the detection algorithm

- Adjusted thresholds specifically for low volume processors

- Set a goal to reduce false positive by 50%

Business solution

- Implemented the dual sev system for different threashold levels

- Helped them build a case for excluding low sev monitoring tickets from their metrics.

The outcome was successful by reducing false postive by 50%

Delegating a task as an EM is a very important skill as it builds confidence in team and help them learn new skills . Below mentioned framwrork I use to delegate task : 

1. Understand the task its impact and its effort. 

  a. If the task can act as a stepping stone for some one will assign hem the task. 

  b. If the task is a repeatable task with medium impact and low effort I will assign this to a junior developer with the ownership and clearly define what needs to be done. 

  c. If the task is is very urgent very impacting and if I have the skill and other are busy with other high priority task then I will do it and share the learning and feedback with the team , else I will assign it to a senior developer d. if a task is low impacting and not urgent then will add it in backlog

e. If that task is very critical and will require a lot of time then will discuss with business and will stop another task and distribute it with the team 

f. Always make sure that thing provides learning and uplevel the skill set for the team.

For Example : 

Situation: Our service availability was low and we need to fix it urgently.

Task : we need to identify the area of improvment reason for down time , 500 error , 400 error  , latencies.

Action : 

1. When I categorized this task this was a high priority and impacting task.

2. I assigned this task to an entry level SDE-3 who wanted to move in lead direction.

3. I set the clear objecting and outcome of this task and asked him to come with a plan and resouces required.

4. I helped him out to set the milestone and success of this task

5. He created a etrics and with the help of an SDE-2 who wanted to work on tech product and SDE-1 who had a now about the system we crated a small task fource that tech lead was leading and I was providing the coaching.

Result : 

1. Tech lead learnied how to own and deliver the product . 

2. Stake holder received weekely communication about the metricies number of 500 bug fixed.

3 .Task force got to know various area of code.

4. And we idenitifed some tech debts that were not immediate actionble but need to solve in long run.

I would consider if this task can be stepping stone for someone to grow into a role, this task would be assigned to them.

If this work is impactful and has visiblity to Executives I would assign it to Senior member who can be owner and i have handsoff approach.

If this is a repeatable with less complexity task I would assign it to a Junior Engineer for them to take care of automation.

I would communicate the reasons of delegation and set clear expectations for the task completion.

For all the tasks I would gauge the current skills of the candidate and provide coaching mentoring  and skill development for them to handle the task.

I would share an example where there was a high impact feature to be built. I engaged a staff engineer for Scoping and architectual design and engaged juniors for coding tasks.The junior members were provided coaching on the translation of architectural design to Service definition.

Delegation is a very important skill. As it shows your team members that you struct them and is genuinely interested in their growth

urgency and importance - I evaluate the urgency of the task and the impact. If the task is a time bound task which has signification impact on the customer then I tend to do it myself to not put the team member on the spot. complexity - If the complexity is high I oversee the task myself. I assess the potential risk of delaying the task to the business and consider the impact on stakeholders before making a decision. expertise and skill level - if the task maps with the skills of the individual I like to delegate to promote development collaboration - delegation can empower team towards autonomy which is good for developing leaders

e.g.

I delegated running weekly RCA call for the platform team to the Architect in my group

• Expertise - In 2 quarters time, architect has build good knowledge about the overall system

• Collaboration - Architect has run major cross team projects and showcased skills needed to collaborate with respect with other teams

• Attention to detail - Architect was part of the meeting for 3 quarters and knew how to drive the post mortem meetings

• Reverse shadow and feedback - I gave the opportunity to the architect to run the RCA call and did reverse shadow and provided feedback later.

• Expectation setting - Setting up the right expectations for running the call and identification of the long term tech items for roadmap,

• Review process - setup a biweekly cadence with architect to review this

Yes, I am an hands on engineering manager

- It all depends on the task and how open ended it is plus which engineers I have eg

If task is open ended all my senior, staff engineers are busy, I will get  clarity on prioritisation eg whats the priority based on eg what benefit it will have? more revenue, customer satisfaction, decreased tech debt etc. Accordingly I will prioritise.

If it requires immediate prioritisation and I cant get my senior folks in this, but I see that this problem can challenge a junior dev, I will try to create clarity for them and assign it to them post clarity.

If none can be done, I will negotiate with senior leadership on prioritisation and compare with all the other items on my teams plate. I will also check if somebody wants to challenge themselves and volunteer for this problem.