Next Environment Quick Start

​

Overview

​

What was migrated from Nexus

Projects and their entities were migrated as a point-in-time copy (Aug-15-2025):

• Your Private projects
• Your Team projects
• Public project

Within projects, the following content was included:

• Agents
• Pipelines
• Toolkits that were used by your agents and pipelines
• Chats and conversations
• Collections
• Artifacts
• Secrets
• Personal Access Tokens

​

What was migrated as a conversion

• Prompts → Agents: All prompts were converted to agents. You can now create agents without a toolkit (a lightweight replacement for prompts) and use variables with them.
• Datasources → Temporary Toolkits: All datasources were temporarily converted to toolkits and remain connected to agents/pipelines where they were previously linked in Nexus.

​

What was not migrated

• Datasets of datasources: Recreate them using the Index Data capability available in supporting toolkits. See the section below: Index data for a toolkit.
• Integrations: Recreate them as Credentials and attach to your Toolkits. See: How to create a credential and attach it to the toolkit.

​

Next steps

1. Create and attach credentials
   • See: How to create a credential and attach it to the toolkit
   • Guides: Create a Credential · Create a Toolkit · Use Credentials
2. Check your Toolkit Configuration
   • Verify tools enabled (e.g., Index data), correct Credential selected, PgVector and Embedding settings; Save, then test from Tool Testing.
   • See: Check your Toolkit Configuration
3. Check your agents and pipelines
   • Open each, resolve any configuration notices, and run one to confirm it executes.
   • See: Check your agents and pipelines
4. Check your conversations
   • Add your agent/pipeline in Chat and send a simple prompt to validate end-to-end.
   • See: Check your conversations · Create a conversation
5. Index data for a toolkit
   • Use Tool Testing → Index Data to recreate datasets; parameters vary by toolkit.
   • See: Index data for a toolkit · Example: Index GitHub data
6. Configure EPAM AI DIAL keys (strongly recommended)
   • Shared LLM models in Next are evaluation-only with daily limits. For sustained or production use, configure your own DIAL keys.
   • See: Configure EPAM AI DIAL Keys · Settings → AI Configuration

---

​

How to create a credential and attach it to the toolkit

Keep it simple: create the credential, then assign it inside your toolkit.

1. Create a credential: Go to Credentials → + Create → pick type (e.g., GitHub, Jira) → fill fields → Save.
2. Attach to a toolkit: Go to Toolkits → create or open a toolkit → in Configuration, choose the Credential you created → Save.

See the detailed guide: Create a Credential and Add It to a Toolkit. For broader concepts, see Use Credentials.

​

Check your Toolkit Configuration

Before using a toolkit (including indexing), verify configuration:

• Tools: Ensure the correct tools are enabled/selected for your use case (e.g., Index data, Create a file, Read file, etc.).
• Credential: Confirm the right Credential is assigned (private vs project scope as needed).
• PgVector: Select/verify PgVector (or your configured vector store) and Embedding model settings.
• Save: Click Save to persist changes before testing.
• Test: Use the Test Settings section to run a quick test and confirm the toolkit works.

​

Check your agents and pipelines

After toolkits are configured:

• Open your agents and pipelines and confirm all attached toolkits are valid (no error banners or misconfiguration notices).
• Resolve any missing credentials or required fields.
• Execute an agent (or run a pipeline) to verify it responds and completes successfully.

​

Check your conversations

Add your working agents, pipelines, and toolkits to a conversation and test end-to-end:

• Open Chat → add/select an Agent or Pipeline as a participant.
• Send a simple instruction and confirm the response.
• For details, see: Creating a Conversation and Adding AI Assistants.

​

Index data for a toolkit

Create your first index using a toolkit that supports indexing: Supported examples: ADO Repos, ADO Wiki/Boards/Test Plans, Bitbucket, GitHub, GitLab, Confluence, Jira, SharePoint, Artifact, Figma, TestRail, Xray Cloud, Zephyr Enterprise, Zephyr Scale. Steps:

1. Open the toolkit and ensure the “Index data” tool is selected/enabled.
2. Go to the Test Settings section → select the “Index Data” tool from dropdown list.
3. Fill the required parameters/options (these vary by toolkit: collection suffix, branch name, blacklist files etc.).
4. Run Tool and wait for completion. Check the output for any errors and adjust configuration if needed.

​

Configure EPAM AI DIAL keys (recommended)

Quick steps:

• Request a key from EPAM AI DIAL (see the guide below).
• In Next env: Settings → AI Configuration → + Create → choose “AI DIAL” → enter API Base, Key, Version → Save.
• Create LLM models: + Create → LLM Model → set exact model name from EPAM catalog, context window, max output tokens → select your DIAL credential → Save.

Full how-to: Configure EPAM AI DIAL Keys

​

Questions or issues? Contact Support