Hr ops · Production

SLB uses Dataiku to reduce well construction tender analysis from 8 hours to 20 minutes

The problem

SLB needed a single data science and AI platform with no- and low-code interfaces to efficiently process domain data at scale. Manual well construction tender analysis took approximately eight hours per well due to unstructured daily drilling report data and unconscious-bias errors. Reservoir pressure analysis took at least one week and lacked any visualization dashboard. The company also faced an annual unplanned employee attrition cost of $80-$200 million with no automated means to identify at-risk employees early.

Workflow diagram · grounded in source
1
Well construction bid received
Trigger
SLB must correctly size its response to a bid to ensure profitability while providing competitive prices to customers.
▾ source quote
“they must correctly size the response to the bid to ensure that the project is profitable while providing competitive prices to customers”
2
Data-driven tender analysis in Dataiku
Ai action
SLB developed a data-driven approach with Dataiku for well construction tender analysis.
▾ source quote
“SLB developed a data-driven approach with Dataiku that has so far been used to assess more than $10 billion worth of well construction tenders and allows engineers to do the same analysis in just 20 minutes”
3
KPI extraction and risk forecasting
Ai action
The workflow extracts key performance indicators, builds an operational sequence, and forecasts risks for each well in the tender.
▾ source quote
“the extraction of key performance indicators before eventually building an operational sequence and forecasting the risks associated with each well of an invitation to tender”
4
Auditable 20-minute analysis output
Output
Engineers complete the well construction tender analysis in just 20 minutes using a structured, auditable, and data-driven process.
▾ source quote
“allows engineers to do the same analysis in just 20 minutes. In addition, the updated process allows for a structured, auditable, and data-driven approach to predicting the time it will take to drill the wells, as defined in the tender's …”
5
Automated pressure detection
Ai action
A Dataiku workflow automates the identification and gathering of stabilized reservoir pressure data.
▾ source quote
“developed a reservoir pressure detection tool that automated the identification and gathering process of stabilized pressure”
6
Pressure results visualization
Output
Results are easily visualized in Spotfire and the monthly pressure trend analysis process is now 76% faster.
▾ source quote
“allows easy visualization of results in Spotfire, and the process to analyze monthly pressure trends well by well is now 76% faster”
7
Employee attrition prediction
Ai action
Dataiku data pipelines analyze salary, vacation, performance, and career stagnation data to identify at-risk employee populations.
▾ source quote
“built data pipelines from troves of data (i.e., across salary information, vacation data, performance and career stagnation information) to notify talent managers across the company about at-risk populations”
8
Talent manager notifications
Output
Talent managers receive notifications about at-risk populations so they can take actions as early as possible.
▾ source quote
“notify talent managers across the company about at-risk populations so they can effectively take actions as early as possible”
Reported outcome

Dataiku reduced well construction tender analysis to 20 minutes and the approach has been used to assess more than $10 billion in tenders.
Reservoir pressure analysis became 76% faster. The HR predictive model enabled SLB to retain $18-45 million annually in employee value, and the Skills2Career dashboard saved hundreds of thousands in recruitment costs.

Reported metrics
Well tender analysis time (before)approximately eight hours
Well tender analysis time (after)20 minutes
Well construction tenders assessedmore than $10 billion
Reservoir pressure analysis speed improvement76% faster
Show all 10 reported metrics
well tender analysis time (before)approximately eight hours
well tender analysis time (after)20 minutes
well construction tenders assessedmore than $10 billion
reservoir pressure analysis speed improvement76% faster
reservoir pressure analysis time (before)at least one week
annual unplanned employee attrition cost$80-$200 million
employee retention value preserved annually$18-45 million
recruitment drive savingshundreds of thousands of dollars
overall business savingssave millions
training time reductionreducing the time invested in training by months and years
Reported stack
DataikuSpotfirePowerBI
◆ Does this fit your context?

Compare to your context

Tell us your scale, team, and constraints. We'll show what changes at your size, what fails at your scale, and whether this case is a fit, needs adaptation, or won't scale to you. Free demo, no signup.

Compare to your context →
~30 seconds · free
Source
https://www.dataiku.com/stories/detail/dataiku-slb/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Dataiku reduced well construction tender analysis to 20 minutes and the approach has been used to assess more than $10 billion in tenders.

What tools did this team use?

Dataiku, Spotfire, PowerBI.

What results were reported?

Well tender analysis time (before): approximately eight hours; Well tender analysis time (after): 20 minutes; Well construction tenders assessed: more than $10 billion; Reservoir pressure analysis speed improvement: 76% faster (source-reported, not independently verified).

How is this hr ops AI workflow structured?

Well construction bid received → Data-driven tender analysis in Dataiku → KPI extraction and risk forecasting → Auditable 20-minute analysis output → Automated pressure detection → Pressure results visualization → Employee attrition prediction → Talent manager notifications.

Related hr ops cases

WHAT TO DO WITH THIS

Now compare it to your context

This case is one data point. Whether its pattern fits you depends on your volumes, your stack, and your exception load — that comparison is the step no case study can do for you.

Browse the Hr ops cases →