Edit

Data Analyst responsibilities

The goal of this page is to clearly outline what is expected from a person in the Data Analyst role. Provide understanding of workflow, artifacts to produce, and cooperation with other roles.

Regardless of the type of Use Case, all Data Analyst responsibilities can be divided into three main stages:

1. Data set collection. The most considerable and important stage makes up more than half of all DA work. Mistakes made and not corrected here may entail serious risks in terms of timelines and achievement of success criteria. As a prerequisite to this stage, DAs should get familiar with different aspects of the project, clearly understand the business logic of the use case, what type is it (Information Extraction/Classification), what fields/classes will be used, document types and formats etc.
2. Model Training and analysis of results. This stage is implemented by close cooperation of DA and ML Engineer.
3. Preparation of the final report. At the end of the project, final results should be presented to the customer.

To identify steps described below with a use case workflow, see the workflow diagram here:

Stage 1. Data set collection

1. Know documents' specific and business logic enough for further data processing.
2. Guarantee an acceptable quality of the data set after OCR conversion.
3. Split documents into batches by some criteria (if not done by the customer).
4. Define tagging logic for each template.
5. Train and qualify SMEs:
   • Hold training session for SMEs
   • Provide clear and specific instructions for labeling
   • Hold Qualification
6. Prepare Manual Tasks for tagging and assign batches to SMEs.
7. Validate labeling results.
8. Split data set on training and test.

Stage 2. Model training and analysis of results

Model training

1. Launch model training.
2. Launch IE/Classification on test set (with assistance of MLE).
3. Run statistics aggregation business process.

Analysis of results

1. Analyze type of mistakes and possible reasons for them.
2. Define rules for post-processing.

ML improvements & retrain

This step is mostly conducted by ML Engineer and Data Scientist. However, Data Analyst can support in analysis of interim results.

Stage 3. Final report preparation

Analyze final model results and prepare the report for the customer. More information about report preparation can be found here.