Elodie Collins 
Seekr has unveiled new SeekrFlow Agents capabilities to enhance control, customization and reliability of artificial intelligence agents when deployed across government and enterprise environments.
The AI company said Wednesday that Structured Output, Custom Tools and Docstring Validation will simplify development and accelerate adoption of agentic AI.
New SeekrFlow Features
Structured Output enables users to define schemas that agents must follow, such as JSON or Pydantic. It aims to address challenges related to the unpredictability of agent output, which can prevent operationalization.
The new feature ensures that agents generate schema-adherent responses that are immediately ready for use.
Meanwhile, Custom Tools allow users to embed proprietary Python functions and connect agents to internal Application Programming Interfaces or systems like Slack and Outlook. Developers can also utilize the tools to use traditional machine learning models with generative capabilities to create agents that can reason and act.
Custom Tools and Docstring Validation enhance customization. A docstring inside a Python describes a function, its parameters and its return values.
The capabilities are now available on SeekrFlow. According to the company, new features, such as Secret Manager and Observability Platform, are under development.
SeekrFlow was recently added to Amazon Web Services’ GovCloud through AWS Marketplace.
