Docs
Build useful managed AI agents with business context.
Practical implementation notes for teams using TactasAI to turn company knowledge, repeated workflows, and connected tools into reviewed outputs and controlled action.
Documentation
Overview
What TactasAI docs cover and how to use them.
Core Concepts
Agents, context, review, actions, and managed improvement.
Launch Model
The review, connect, launch, and improve path.
Quickstart
Choose one workflow, prepare context, and launch a supervised TactasAI agent.
Data Readiness
Prepare source files, records, and business rules so answers stay grounded.
Agent Design
Define roles, approval points, outputs, escalation logic, and boundaries.
Integrations
Map business tools into readable sources, write destinations, and controlled actions.
Source Connections
Plan the sources an agent can read for each workflow.
Controlled Actions
Design action paths that require the right review and confirmation.
Operations
Review quality, measure usage, handle exceptions, and improve deployed agents.
Quality Review
Use reviewer feedback to improve prompts, context, and outputs.
Launch Checklist
Confirm ownership, data, approvals, and rollout readiness.
API Overview
Base URL, authentication, request headers, retries, and supported API areas.
Go SDK
Install and configure the official Go client for the TactasAI v1 API.
Agent Runs
Create sessions, submit messages, watch run events, and handle approvals.
Knowledge Search
Search workspace knowledge and manage source sync with the API.
Errors And Retries
Handle non-2xx API errors, request IDs, idempotency, and retry behavior.