Begin by familiarizing yourself with the Zendesk Talk API. This API allows you to access call data such as call details, recordings, and other related information. Review the API documentation to understand the endpoints available, authentication methods, and data formats (usually JSON).

Ensure you have an AWS account with permissions to create and manage DynamoDB tables. Navigate to the AWS Management Console, and create a new DynamoDB table. Define the primary key structure based on how you intend to query the data (e.g., Call ID as the primary key).

Create an API token in Zendesk for secure access. Go to the Zendesk Admin Center, navigate to the API section, and create a new token. Store the token securely; you will need it to authenticate your API requests using basic authentication (email/token).

Write a script in a language like Python to extract data from Zendesk Talk using its API. Use libraries such as `requests` to handle HTTP requests. Fetch the required data by calling the appropriate API endpoints. Handle pagination if necessary to retrieve all records.

Once the data is extracted, transform it into a format suitable for insertion into DynamoDB. Ensure data types match those supported by DynamoDB (e.g., strings, numbers, booleans). Consider data structure alignment with your DynamoDB table schema.

Use the AWS SDK for your chosen programming language (e.g., Boto3 for Python) to write the transformed data into DynamoDB. Implement a function that inserts each record into your table. Consider batching requests to improve efficiency, as DynamoDB supports batch writes.

Finally, automate the data transfer process to keep your DynamoDB table updated. Use AWS Lambda to execute your script at regular intervals, or set up a cron job on a server if you prefer. Ensure error handling and logging are implemented to monitor and debug the process.

By following these steps, you can effectively transfer data from Zendesk Talk to DynamoDB without relying on third-party connectors or integrations.