Here are the functions DC Agentiqs provides.
Chat
A chat feature that comes standard and delivers the following.
- Freely choose the LLM model
- Adjust the thinking level of reasoning models to suit the task
- Talk with specialized AI agents
- Knowledge search powered by RAG
- Reference current information through web search
- Get evidence-based answers with the code interpreter
- Attach files by drag and drop
- Long-term memory through context enhancement settings
- HITL (Human In The Loop) lets a person check the AI's work
- Check context usage and execution cost
Agent Design
Build the agent you need yourself, using the following structure.
| System Prompt |
Defines the agent's role, purpose, work steps, output format, and more. |
|---|---|
| Tools | Select the tools the agent integrates with and their functions. |
| Knowledge | Select the documents the agent uses as reference information. |
| Subagents | Configure calls to other agents. |
| Skills | Configure reusable definition sets containing business rules, specialized knowledge, scripts, and reference materials. |
Knowledge Creation
Covers everything you need, from building RAG to running it in production.
| Build RAG | Build RAG by specifying the path to local files. |
|---|---|
| Search settings | Set the maximum number of search results and the relevance threshold. |
| Index updates | When documents are updated, differential updates keep the index current. |
| Document conversion | Convert Excel grid sheets, ruled PDF tables, shapes, images, and sequence diagrams into information AI can use. |
Tool Integration
A standard set of tools that agents can use, ready out of the box.
Standard tools
- MS Office products
- File access
- Command execution
Design toolsProfessional Edition only
- Next Design
- Simulink
Management tools
- GitHub
- Confluence
- Jira
- TimeTracker
- Lightning Review
MCP Integration
- Connect to any MCP server.
WorkflowProfessional Edition only
Automate work by calling agents, tools, and knowledge appropriately using the following elements.
- Control flow
Loop, conditional branching, parallel, sequential, while
- Integration & approval
Run Python scripts
- AI processing
Agent calls, AI-based branching, knowledge search, LLM calls
- Other
Tool calls
- Schedule execution
Run workflows on a specified schedule to automate recurring development work
- Debugging
Investigate problems using breakpoints, step execution, and variable inspection and editing
- Version history
Save edits as named versions and restore a previous state when needed
AI-assisted creation of agents and workflows
- AI creates the agent.
- AI creates the workflow.
Workspace
Manage agents, tools, knowledge, and workflows together as a Workspace, in whatever grouping you choose.
- Sharing
Manage agents, tools, knowledge, and workflows together as a team.
- Moving
Move agents and knowledge from one Workspace to another.
- Duplication
Duplicate agents and workflows.
- Shared organization settings
Distribute settings such as AI provider configurations in encrypted form and keep member settings up to date.
Environment & Usability
- Supports Japanese and English interfaces, so it works smoothly in global environments too.
- Supports local LLMs through Ollama and LM Studio.
- Connects to Azure OpenAI and Microsoft Foundry using Entra ID authentication.
