IBMhadeepq

What's my job?

I'm definately in the Data Science area, but please take a look. It feels that I am all in 1 and it's draining. Should all this DS work feel like this? Do you also do the entire work from start to end? - I am doing the architecture and consider all the cases - I create all the tickets in jira - I follow the workflow - I write documentation - I do testing for my code - I make the data pipelines (ingestion, cleaning, etc) - I explain to the TAM what they should say to upper management - I explain to the business analyst what we are doing and what is the workflow - I explain to my dearest colleagues the architectural decisions, why I chose to do this or that, and what are the next steps... all the next steps.. - I am doing the entire AI model - I am deploying both data and AI parts to cloud - now they want to have RL like OpenAI and Netflix say what 😣 Is this a normal Data Scientist job? I feel that I also do the DS Architect, Tester, ML Engineer work, a kind of talking manager...... #tech #datascience #softwareengineer #architect #netflix #palantir #databricks #google #apple #amazon #aws #microsoft #openai #cisco #nvidia ✨ please help, you are the experts in tech

New
dogfoodie Jan 15

Sounds like a startup.

IBM hadeepq OP Jan 15

Are you currently working at a start-up rn? Is it the same atmosphere?

New
dogfoodie Jan 15

Yes you wear all the hats in a startup because there's no one to pick up the slack.

Google x-mas 🎄 Jan 15

I do the following, if that helps: 1. Scope the project: what we should be doing, use cases, risk, data availability, timelines, key stakeholders etc. 2. Create tickets 3. Write or review codes or perform analysis: data prep and model 4. Create documentation: Create any slides, write detailed notes 5. Explain to the xfn partners: Key insights, metrics movements, next steps 6. Close ticket 7. Quantify impact of the project, send success emails or do post mortem. What I don't do: 1. Data pipelines 2. Model deployment Engineers help with these tasks.

IBM hadeepq OP Jan 15

Nice! If you don't have to argue with collegues it is perfect. May I ask what level are you? And also if you are in product, consulting or research?

Google tc--; Jan 15

Wow, my turn: 1. Read data from a protobuf 2. Write the same data to a different protobuf

Cisco bellspalsy Jan 18

It sounds like you have a useless team depending on you. You can either leverage that to get more money and a nice promotion or go somewhere else where you get paid more for doing one job instead of 3.

IBM hadeepq OP Jan 18

Thank you!! I was thinking the same. The management don't even understand what they should follow in the meetings or how much each part of the end-to-end architecture take in order to be set for production...

CrowdStrike HXzr18 Jan 18

Dude don’t waste our time unless you post your TC

IBM hadeepq OP Jan 19

39k $ EMEA