Note: this is an illustrative example, based on typical data analysis projects. The data is representative and the company has been anonymised.
A distribution company replaced manual report production with conversational analysis of its data, using EngiAnalytics. Management questions that previously required a request to the team and a wait of one to two days now get an answer in seconds, phrased in natural language by whoever needs the information. The most relevant effect was not saving reporting time, it was changing how often the company looks at its numbers.
The context
The company produced recurring reports in spreadsheets, fed manually from the ERP and other sources. Whenever management needed a new analysis, for example margin by customer or the trend of a category, it had to ask whoever mastered the spreadsheets, who would prepare the cross-analysis and return the result later. Urgent questions ran into this wait.
The challenge
- Dependence on a few people: the knowledge of how to obtain each figure was concentrated in a handful of people.
- Decision latency: the answer took long enough that many questions were never even asked.
- Low-value work: skilled time spent building reports instead of interpreting them.
- Data spread across sources: relevant information was split across systems.
The solution
Engibots implemented EngiAnalytics connected to the company's data sources (see what EngiAnalytics is).
- Connection and definitions: connection to the ERP and other sources, with prior agreement on the definitions of key indicators, to ensure consistent answers.
- Natural language questions: business users started asking directly, without intermediaries, and getting the answer with the appropriate visualisation.
- Security and compliance: access in line with permissions and data handling aligned with GDPR (see GDPR and artificial intelligence).
Results
Illustrative results of this approach:
- Response time: from one to two days down to seconds, for routine management questions.
- Autonomy: management and sales teams no longer depend on third parties to get information.
- Frequency of use: with instant answers, the company started consulting its data much more regularly.
- Skilled time freed up: whoever used to produce reports now focuses on analysis.
Lessons
- Speed changes behaviour: when the answer is instant, more questions get asked and better decisions get made.
- Definitions are the foundation: the value of conversational analysis depends on terms like "margin" or "customer" having a single definition (see conversational analysis vs dashboards).
- Analysis and automation reinforce each other: data generated by automated processes becomes immediately available for analysis, closing the loop between EngiMatrix and EngiAnalytics.