Begin by familiarizing yourself with the Zendesk Talk API, which is the primary tool for extracting data. Review the API documentation provided by Zendesk to understand the available endpoints, authentication requirements, and rate limits. This will help you identify the specific data you need to extract, such as call records or voicemails.

To access Zendesk Talk data, you need to authenticate your requests. Set up an API token in your Zendesk account by navigating to the Admin Center, selecting "API" under "Channels," and then creating a new API token. Use this token for authentication by including it in the request header when making API calls.

Write a script in a language like Python to request data from the Zendesk Talk API. Use libraries such as `requests` to send HTTP GET requests to the required endpoints. For example, you might request call records by sending a GET request to `https://yoursubdomain.zendesk.com/api/v2/channels/voice/calls.json`. Parse the JSON response to extract the data fields you need.

Ensure that your PostgreSQL database is set up and accessible. Create a new table or tables to store the Zendesk Talk data, ensuring that the table structure matches the data format you plan to extract. Define appropriate data types for each field, such as `VARCHAR` for text and `TIMESTAMP` for dates.

Process the extracted data to fit the PostgreSQL schema. This might involve transforming date formats, cleaning text fields, or normalizing data values. Use a scripting language like Python to automate this process, ensuring that the data is clean and consistent before loading it into the database.

Employ database libraries such as `psycopg2` in Python to connect to your PostgreSQL database and insert the transformed data. Construct SQL INSERT statements for each record and execute them using a cursor. Be mindful of batch processing to handle large volumes of data efficiently and to manage database transaction limits.

To keep your PostgreSQL database updated with the latest Zendesk Talk data, automate the extraction and loading process. Use tools like `cron` on Unix-based systems or Task Scheduler on Windows to run your script at regular intervals. Ensure your script handles exceptions and logs errors to facilitate troubleshooting and ensure data integrity. By following these steps, you can effectively move data from Zendesk Talk to a PostgreSQL database without relying on third-party connectors or integrations.