TL;DR: My company does not assign me data scientist tasks and I am feeling like I am being wasted.

I am currently working in a consultancy firm as a data scientist (four years experience). In general, my work consisted in coding proof of concepts for clients who wanted to check if a problem they wanted to solve with AI is viable. I always tried to do my best and not only exercise as a data scientist, but also I've always strived for improving my software engineering skills and deeply understand concepts as technical debt, testing, CI/CD, etc.

I always have had the initiative to learn myself about packaging my code to make it available, while practically 100% of other data scientists I've met only worry about their notebooks. The thing is that the company I work for, just like other typical consulting firms, is very fast paced and sometimes there are project needs that can't be met with the current available workforce.

For some projects, my managers have given me the responsibility of several other things not really related to data science, like mentoring, integration of other data scientists' code, debugging, testing, client presentations, cloud deployment and integration, etc.. which I would not mind if they made everyone work on this.

The problem is that they are only making me do this, because I am a quick learner and I put much effort in my work, so I can get things done even if I am a newbie (e.g. cloud stuff), and other less skilled data scientists are always in their bubble, working with their notebooks.

It has reached to a point in which I have not done data science for an entire two months, and I would not mind if it was a one-time thing, but it is getting more and more constant. The worst feeling I have of everything is that I feel like I am being wasted because I consider myself a good data scientist.

Is all of this normal? Is there anything I could do, like approaching my superior about it?

  • 1
    Comments are not for extended discussion; this conversation has been moved to chat.
    – motosubatsu
    Commented Oct 27, 2021 at 15:25
  • 1
    Q: Does your company have data scientist tasks that they could be assigning to you? Commented Oct 28, 2021 at 16:31
  • 4
    Do you enjoy your job as it stands?
    – Caius Jard
    Commented Oct 28, 2021 at 16:46

10 Answers 10


Let me rephrase the question: Is it normal to be handed other tasks than initially envisioned when you have shown you are capable to handle them?

Yes, absolutely! There is a problem, you are able to solve it, and off you go!

Data scientists often don't see themselves as programmers or engineers, and don't want to be bothered by this. So a good solution is to hire explicit engineers, who handle stuff like packaging, and who support the scientists in writing tests (where it makes sense). But because a lot of data scientists companies don't understand this, they simply hire only data scientists, and then data scientists are forced to do stuff they don't like.

Some companies implement a rotation system where everybody has to do unloved duties sometimes. Others hire explicit people to handle them. If unloved equals low prestige, it's a sure recipe to burn through people fast. If unloved work is still appreciated, that's a much better situation.

What you can do: Ask yourself, what are you willing to do? Would you change companies over this? Would you accept this new role for a payrise, better title, and other perks?

Then talk with your superior. Make your superior aware that you don't see this as long-term solution. Present the options you see for improving. If you are willing to leave, don't threaten to leave. Just make very clear you won’t stand for nothing changing at all.

You have to be patient, because changing things like this takes time. On the other hand, don't let them lead you on forever. If you don't insist on change, it's convenient for them to keep you in that role. So if you talk to your superior, talk about timelines.

After seeing the comments: Some things you describe sound like an adjacent role (e.g., CI/CD), while other things like mentoring and client presentations indeed sound like advancement. Make sure to distinguish those two in your head.

  • I perfectly understand what you're explaining here. The thing that happens in the company I work for is usually the same though: There's a deadline for the client next week, we have a sh**load stuff still to be done (e.g. full deployment and integration with a platform), and I am always assigned these tasks, I do not know why. Maybe because there is other people but they would not be able to accomplish these tasks in that amount of time, but that is completely unfair in my opinion. Is like being a good learner and hard-working is ironically rewarded with tasks I don't like to do Commented Oct 26, 2021 at 9:07
  • 30
    yes, good work is often punished with more work. Google for performance punishment or something similiar. e.g.: amanet.org/articles/…
    – Benjamin
    Commented Oct 26, 2021 at 11:57
  • 2
    @IvánSánchez Absolutely, I'm in a similar position myself except it's the ML/AI stuff I don't want to be doing (and even less so, coding at the production level). It is changing though, so what you need to recognize is that there is a conflict of interest. Your management values you as a fixer, the guy who actually gets stuff done. The pinnacle of this career might be a CTO - it is a viable spot to be in but not necessarily what you want. So either push back on it and refuse some work or ask for extra resources. Your company burdening you with increasing amounts of work is a form of tech debt.
    – Lodinn
    Commented Oct 27, 2021 at 2:20
  • "Performance punishment" is insidious because it implicitly happens almost everywhere. If you are good at solving hard problems, you'll get more of them. If you are fast, you are going to get more tasks assigned. If you have skills that others don't have, you get tasks that require these skills (even if you prefer other tasks). Some companies are more aware of this problem than others (and try to rectify with compensation etc.), but I have yet to see a place where this didn't happen at least on some level.
    – xLeitix
    Commented Oct 27, 2021 at 10:59
  • 3
    Honestly, long-term the best approach is to accept that competence will be rewarded with more work, to be proud of your own competence, and to fight that your competence is acknowledged and rewarded appropriately (money, PTO, goodies, a pad on the back, whatever works for you). Because the situation per se isn't going to change.
    – xLeitix
    Commented Oct 27, 2021 at 11:04

Is all of this normal? Is there anything I could do, like approaching my superior about it?

It's not unusual for people to be assigned some tasks that need to be done, even if it is outside the main portion of their role. Sometimes, someone just needs to do them. In this case, they chose you.

That said, if you are spending more time outside of your data scientist role than you would like, talk with your boss about it.

And if it looks as if this will be an ongoing thing that you don't want, find a new job that will let you concentrate on what you prefer doing, and put this one behind you.

  • 1
    Quite… and what's actually wrong with the "extra" roles? Going (slightly) sideways, isn't one of the most common complaints about management in technical areas, that the managers don't understand the technicalities? Commented Oct 27, 2021 at 20:30

I think there's a lot going on here. I'm taking from your message that you see yourself in a negative place, however I believe you need to reframe that thought. There's a few questions you need to ask yourself before speaking to your manager:

  1. Is consulting the right industry for how I want to work?
  2. Are my expectations of my work reasonable with the organization I'm working in?
  3. Are my observations of my DS colleagues based in fact or judgement?
  4. Who do I want to be in the work place?

Consulting, by its nature is a fast-paced all-hands-on-deck industry. Principals will rely heavily on can-do and will-do people and others will eventually get weeded out. Only the cream that rises to the top will be saved and this doesn't always mean the most talented individual, but the one that consistently delivers good quality work for the client.

Looking at the work you are being given, to me it appears a split between supervisory work (mentoring, integration of other Data Scientists code), trusted associate work (debugging, testing, client presentations) and other technical task you're capable of (cloud deployment and integration). My impression is that you are being relied on to do this stuff because you are trusted and capable, and you're not another DS who hides away in a corner plugging away at MatLab, Python, Apache, Excel, etc.....

The next thing to consider is your particular organization. Is the consultancy known as an "everyone chips in" or a strict "everyone sticks to their lane" type of place? If it's "everyone chips in", then it's the other DS's who aren't toeing the line and will see the consequences, not you. Even if you go out alone and hang a freelance shingle for yourself, you'll end up doing a lot of ancillary tasks for clients that have little or nothing to do with DS.

Another consideration is if the consultancy runs its own DS practice? If so you should be speaking to the practice lead about getting further DS tasks. Consultancies love an associate who can manage and organize a busy workload and won't disrupt something that is working well. To change this you need to be assertive and should maintain an open dialogue with the practice lead, understand how they see things and review your time recorded with them to show how much time is spent on other tasks. The person who is doing the assignments needs to know how you feel.

In terms of your DS colleagues, you shouldn't be too concerned with what they are doing and how they are doing it. Unless you're across their work you likely don't know the full story. While difficult, it's best to maintain an objective view that's not full of judgement in the workplace.

Finally, the biggest question you need to ask yourself is, who do I want to be in the workplace? This is something only you can answer. Do you want to progress into a management role, or do you want to be a pure data scientist? This is a very hard question to answer, especially at this stage of your career. If I'm reading things correctly, it would seem that the organization trusts and appreciates you so this is likely to be something that is put in front of you sooner rather than later. So it's up to you to understand whether the work you are doing is fulfilling for you; if it is then why try to change it, if it isn't then it's time to act before data science gets further out of reach.

As background, I'm a financial analyst who has yo-yoed my whole career between management and senior advisor roles. I enjoy being an analyst far too much and get bored of supervisory/management very quickly. I often fall into management roles because I have a lot of leadership competencies, but I'd rather be a thought leader in the areas I enjoy working in, than a people leader. My first management opportunities appeared about 4 or 5 years into my career and I thought I wanted it at the time. It was very good experience for me, but over time I naturally drifted back to the numbers. I've worked in consultancies, been a freelance consultant and been employed directly in very large organizations.


The answer to this question is a question. What do you want to do in your career going forward?

In three, four, five years, are you hoping to be a senior data scientist, doing data science (just, better and for more money)? Or would you be interested in a different path - management?

What you're doing now is heading down that second path. Learning more about the organization in all the different parts, helping out where needed, mentoring, client presentations - that's all stuff that will get you on the path to get into management.

It may well be that this is the idea - you might want to ask your manager if that's what they're thinking.

If that's what you are interested in, then just make sure you understand the plan, and understand what the process will be. If it's not, if you just want to stay a Data Scientist and do Data Science, then it may be that this isn't the right fit for you - the company may not have a need for a full time Data Scientist, and you may want to find a company that does.


Ask what you need to do for a promotion, because you're already working at a higher level.

Simply put, it sounds like you're working on the level of a team lead or senior engineer. The fact you're mentoring alone is demonstrating your ability to perform one of the key tasks of a senior software engineer, and some of the other tasks you're handling like client presentations and integrating other engineer's code might be leaning that way as well.

So, if you're not already a senior engineer, I would say that it might be time to schedule a meeting with your boss to discuss that you are operating on a more senior level than your colleagues, what the roadmap to being promoted to a more senior level would look like, and what skills you would need to demonstrate that you haven't already demonstrated.


Is all of this normal?

Yes, most especially it's typical of companies that don't have a ton of experience with using Data Science, but sometimes even companies with a lot of experience in the arena still find more use out of a generalist than they would a specialist.

Is there anything I could do, like approaching my superior about it?

Yes. In general it should always be possible to broach the concept of a mismatch between your expectations and the work that's being asked of you. Assuming a functional organization, and professionalism and open mindedness when discussing the issue.

This is what supervisors and managers get paid to do, and as much as that may seem like a unidirectional conversation sometimes, this is not always the case.

By all means, interrogate why this stuff is ending up on your plate, and make your preferences known, but it would be wise to start this conversation with a few baseline expectations:

  • You will need to compromise
  • You may be called upon to suggest a reasonable alternative
  • What someone else considers reasonable may not align with what you consider to be reasonable
  • There are reasons that things are the way they are now, and they seemed like legitimate reasons to the people making the decisions at the time they were made
  • Any change will take time

If you stay honestly committed to those principles and open to new ideas, you can turn almost any legitimate workplace complaint into a constructive discussion about how to solve a problem.


Short answer to your question: yes! You know enough in a vast amount of areas to be utilized, especially when there is a need. Since data science is highly interdisciplinary, this is the case for most of data scientist, they need to have a fundamental level of knowledge in a lot of related topics. And since the knowledge is there, it is tempting for employers to use this already existing knowledge.

To stick with the analogy of a Swiss army knife: when properly building a house from scratch, you would not use the army knife's screwdriver, for this use case you would probably buy one or multiple professional ones. However, if you want to build a shelf, you might consider it. Even though it's not the ideal tool, you can achieve your goal with some minor inconveniences. Buying a new screwdriver just for the shelf could be considered "waste", if you don't plan to build a series of shelves for each and every room. Plus you don't need to buy a separate bottle opener.

Within a consultancy, you often have short-term need of people for specific roles and while you may not be the ideal candidate, you showed that you do jobs outside of your primary area well enough to not hire somebody made for this role. If you'd be at a tech company, things might look different, but here it's about selling a certain level proficiency in a topic to customers in need, so often enough that extra mile by having an expert in this field is not even necessary. Billable hours are worth more than waiting for the ideal project for each candidate.

Unfortunately, I work in this area and I know quite a few people that were in your exact situation (Data Scientist doing cloud engineering tasks or building DevOps pipelines), eventually unhappy and some of them left their jobs, so to answer your question: I don't think it's unusual. Your best shot is to address you are unhappy with your current tasks. If they want to keep you, they will change it, in case they do not change your assigned tasks, they should not keep you.


I think this is normal and acceptable within reason.

Every project will have different needs. I am also a data scientist working on the algorithms team for a contract. There is only one other person on my team and she is an SWE by title. There's just too much SW work for her to do alone, and so I help her out. I'd say 50% of my time is spent on software engineering: unit tests, building out frameworks to carry my models into production, etc. I'm even doing 99% of my work in Java, not Python or R.

I don't mind because I am diversifying my skillset which makes me more hireable. I also know that I can move to another project or contract given enough time spent on this one.

So, my advice to you is:

  1. soak up these SWE skills
  2. work on your DS on your own time (you should probably be doing this anyway)
  3. Put in 6-12 months on your current project, and then ask to be moved to a new project/client/account.
  4. If you can't move, quietly start your job search. And be happy you can now put SWE expertise onto your resume.

By your comments you are the only person capable of doing the work; and incidentally you are also very good at it.

There is a saying, "love it, change it, or leave it", which applies to all problems and gives a good framework to come up with ideas:

  • Love it: learn to love your new role. You are an incredibly rare specimen, one of the most sought-after people in the whole IT field, as of 2021. Either simply enjoy the complexity of what you do, maybe even dig in deeper, or use it as an advantage in your next salary hike. The field of Data Science will not go away any time soon, as long as you keep a handle on what's going on there, it will be fine, career-wise.
  • Change it: you commented somewhere that there is no other person capable of doing your unwanted tasks. So lobby for employing some DevOps Engineer, a CI/CD expert, an Ops person or something along these lines - their CV should contain the things you want to hand off to them.
  • Leave it: polish your CV and find something else. This time around, make sure your new employer understands why you are leaving, and that it would be unacceptable for you to land in the same predicament again.

The problem is that they are only making me do this, because I am a quick learner and I put much effort in my work, so I can get things done even if I am a newbie (e.g. cloud stuff), and other less skilled data scientists are always in their bubble, working with their notebooks.

One of signs of a poorly run companies that good employees as reward for good work get more work without being noticeably more compensated than their less productive peers. So if your salary is not matching your productivity that might be a bigger problem for you than not doing the data science.

As for "is it normal": it is common that physicists/mathematicians/data scientists end up doing "regular SW stuff" so it is up to you if you are ok with that.

If not talk to your boss about your "growth direction" and explain to him that you want to do more data science so you can grow in that direction.

If they refuse you could always search for another job, but make sure to not ruin your current work situation by looking too arrogant/entitled.

Not the answer you're looking for? Browse other questions tagged .