Data Engineer
You extend the platform where configuration stops. A Data Analyst models what already fits the semantic layer; you build what does not — custom Python logic, computed classifications, automated pipelines, and insights that write themselves. Athena's proactive surfaces and a customer's automated reporting all trace back to a task or a flow authored by a Data Engineer.
The craft is in reaching past the platform's built-in surfaces when they cannot express what the business needs, and doing it safely — versioned, scheduled, and observable, rather than run once and forgotten.
The surfaces you will spend the most time in:
- Notebook under AI Workbench for Python and the
conversightandcsSDKlibraries - Smart Analytics for publishing computed tables back into the semantic layer
- Tasks and Flows for orchestration, scheduling, and event triggers
- AI Insights and Briefings for surfacing what you have built
Prerequisites
Confirm the following are in place before starting the learning sequence:
| Requirement | Detail |
|---|---|
| ConverSight account | Provisioned with the Data Admin role, with Notebook access under AI Workbench. Confirm this under Settings → Profile before Step 1. |
| Workspace URL | The web address of your organization's ConverSight workspace (for example, https://<yourorg>.conversight.ai). Confirm this with your administrator if it is not already bookmarked. |
| Login credentials | Either SSO through your organization's identity provider, or a ConverSight username and password. Multi-factor authentication may be required depending on workspace settings. |
| Internet connection | A stable connection throughout — ConverSight runs in the browser. |
| Supported browser | A current version of Chrome, Edge, Firefox, or Safari is recommended. |
| Python fluency | Comfortable working in Python — Pandas at minimum, and Polars is a strong advantage on larger datasets. |
Learning Objective
By completing this course, you will learn how to:
- Access the ConverSight platform securely as a Data Engineer and confirm your role and permissions.
- Work in the ConverSight Notebook and choose correctly between the
conversightlibrary andcsSDK. - Perform complex data manipulation in Python that the semantic layer and Data Explorer cannot express.
- Build Smart Analytics tables from computed results and publish them into the semantic layer.
- Author, register, and promote Tasks, and compose, schedule, and trigger Flows.
- Create AI Insights (Proactive Insights) from code, scoped correctly by role.
- Write to V2 object storage and generate and publish the resulting metadata.
Learning Sequence
Five steps. Each pairs a how-to to Read, the work to Do, and a check to Verify before you move on.
Step 1 of 5
Sign In and Orient
Sign in · Notebook
The notebook session opens on either the Standard or Machine Learning base image.
Snag: If Notebook is missing under AI Workbench, the role is not yet Data Admin. Ask an org owner to assign it.
At a Glance
| Certification | Hours | Knowledge Check | Practicum | Pass Mark |
|---|---|---|---|---|
| Data Engineer | ~12 | 90 min | 7 days | 70 / 100 combined |
Coverage
Nine competencies, weighted across the knowledge check (40%) and the practicum (60%):
| # | Competency | Knowledge | Practicum |
|---|---|---|---|
| 1 | Notebook environment, session limits, base images | 3% | 3% |
| 2 | conversight vs. csSDK: connection, metadata, query styles | 5% | 5% |
| 3 | Python manipulation, and the Python vs. semantic-layer boundary | 5% | 9% |
| 4 | Smart Analytics: create, overwrite vs. append, publish, delete | 5% | 8% |
| 5 | Task authoring, registration arguments, promotion, versioning | 7% | 10% |
| 6 | Flow composition, parameters, registration | 5% | 8% |
| 7 | Scheduling, event triggers, notifications, run status and logs | 5% | 8% |
| 8 | Proactive Insights from code, roles and groups scoping | 3% | 6% |
| 9 | V2 object storage and metadata publishing | 2% | 3% |