Related Posts
Where do UITs fit in?
Additional Posts in Data & Analytics Consultants
What’s EY DnA tech stack? EY
Had a few questions about Quantum Black / BCG Gamma.
1. Do QB Data Scientists, also have schedules of waking up at 4, reaching the client at 9 and working until 11 PM? How do you even write code when you're sleep deprived ?
2. Do you use MacBooks or do you also use Windows like normal consultants
3. Can people with degrees in Business Analytics actually get into QuantumBlack ?
4. How is the work of a data scientist in tech different? (apart from new projects each 3 months)
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
Not willing to spend the money to set up a program which requires continued maintenance. Most of them think it’s a one and done deal and that everything can be offshored.
Mentor
Extremely generic question. So looking at it from Analytics and DS perspective.
Answers so far already cover lack of Management interest and willingness.
One of the major challenge is that fully automating DS stacks are extremely difficult. Majority of organisations from large ones to cloud hosted unicorns are still trying to converge on the right architecture to allow faster experimentation and deployment. It’s a tough problem to solve with evolving architectures and even more challenging with legacy ones.
It's automated just enough to meet ad hoc demands with a few scattered late nights. But if you want to talk efficiency, it's moral hazard from organizational bloat and bureaucracy.
Step by step...
1. The data science team is happy to very slowly automate internally and not do procurement, because corporate procurement is a pain enough to be its own job and is slooooooooooow. There are ad hoc requests to meet, which is why they were hired as data detectives in the first place.
2. Management is happy the team is filling ad hoc work, is showing incremental results, and doesn't need to deal with budget changes or do layoffs as a result. Cost center budgets don't come back once lost, so if it ain't broken...
3. Upper management is buzzword-happy enough by just having their own profitable data science team, and can say they don't have extra vendor costs.
4. The cycle is perpetuated by salespeople with quotas to meet for next quarter, who aren't incentivized nor qualified to play the long game convincing data scientists to do procurement. Instead they go after management with budgetary control, who as seen in (2) and (3) doesn't care.
All of it is manual because I’m always doing custom adhoc projects than an automated process wouldn’t provide good insight into
Majority is still done manually because consultant’s can’t practice what they preach
If you automate all your work then what does your team do?
If the point is you’re doing a lot of bs/low value work, that’s another point all together.
Too much organizational red tape 🤬
A ton of it is automated. But automation isn’t perfect. In real life, automation fucks up all the time. A lot of our work is smoothing and refining. We let the machines do the boring work while we supervise.