Documentation

Creating Your First Agent

After signing up, create your first AI agent from the dashboard. Go to Agents → Create New Agent. Choose a name, pick an LLM provider (Claude, GPT, or DeepSeek), and set a system prompt that defines your agent's personality and scope. You can refine the model and temperature later.

Next, add a knowledge base so the agent can answer from your docs and FAQs. Then configure channels (chat widget, WhatsApp, etc.) and use the test panel to try conversations before going live.

Understanding Conversations

Each chat with a user is a conversation. Conversations have a unique ID, are tied to an agent, and keep full message history. You can view and search them in the dashboard, export for analysis, and use the API to start or continue conversations programmatically.

Conversations support metadata (e.g. channel, user ID) for filtering and analytics. Escalation rules can hand off to a human when needed.

Dashboard Overview

The SolveoAI dashboard is your control center. From the sidebar you can access Agents (create and edit AI agents), Conversations (inbox and history), Knowledge Base (documents and FAQs), Data Sources (websites, uploads), and Deploy (chat widget, WhatsApp, API). Settings and API keys live under your profile.

Use the getting-started checklist on the home tab to complete setup, and check analytics for usage and performance.

Agent Builder Basics

The agent builder lets you configure your AI agent without code. Set the system prompt (instructions and tone), choose the model and temperature, and attach a knowledge base. You can switch models per agent and use multiple providers (Claude, OpenAI, DeepSeek) from the same account.

Advanced options include response length, suggested replies, and escalation rules so certain queries are handed to your team.

Conversation Templates

Start from a template for support, sales, or FAQ-style agents. Templates come with a pre-written system prompt and suggested settings. You can customize everything and save your own configurations as a starting point for new agents.

Training Your Agent

Train your agent by adding a knowledge base: upload PDFs, DOCX, or text, add FAQ entries, or connect a website or help center. The platform chunks and indexes your content so the agent can retrieve relevant answers. Keep docs up to date by re-uploading or syncing data sources.

Multi-Model Setup

SolveoAI supports multiple LLM providers. In Settings → API Keys, add keys for Anthropic (Claude), OpenAI (GPT), and DeepSeek. When creating or editing an agent, choose the provider and model (e.g. Claude Sonnet, GPT-4o, DeepSeek-V3). You can run different agents on different models and switch anytime.

See the Quick Start and your dashboard for the full list of supported models.

A/B Testing Agents

Compare agents by creating two (or more) agents with different models or prompts, then routing a share of traffic to each. Use the same knowledge base and measure resolution rate, satisfaction, and latency in analytics to choose the best configuration.

Integration Overview

SolveoAI connects to the channels and tools you already use: embed a chat widget on your site, connect WhatsApp Business, and call the REST API for custom apps. Webhooks let you send events to your systems. More integrations (Slack, email, etc.) can be added via the API or upcoming built-in connectors.

Slack Integration

Deploy your agent to Slack by connecting your workspace in Deploy → Integrations. Once connected, you can invite the agent to channels or use it in DMs. Messages in the channel are handled by the agent using your knowledge base; you can configure escalation to human teammates when needed.

Email Integration

Route support emails to your agent by configuring an inbound email address in the dashboard. The agent can draft replies based on your knowledge base; you can review and send, or set up auto-send for simple cases. Outbound emails use your verified domain.

WhatsApp Business

Connect WhatsApp Business via Deploy → WhatsApp. You’ll need a Meta Business account and WhatsApp Business API access. After linking your number and verifying, conversations from WhatsApp are handled by your agent. Configure greeting messages and quick replies in the same section.

Webhook Setup

Configure outbound webhooks in Settings → Webhooks. When events occur (e.g. new conversation, message received, escalation), we send a POST request to your URL with a JSON payload. Sign requests with the secret you set to verify authenticity. Use webhooks to sync data to your CRM or trigger internal workflows.

Custom Integrations

Use the REST API to build custom integrations: create conversations, send messages, list history, and manage agents and knowledge base. Authenticate with an API key (Bearer token). See the API Reference for endpoints and examples.

Authentication

API requests use Bearer token authentication. Create and copy an API key from Dashboard → Settings → API Keys. Send it in the Authorization: Bearer YOUR_API_KEY header. Keep keys secret and rotate them if exposed.

For full endpoint details and examples, see the API Reference.

GraphQL API

A GraphQL API is available on Pro+ plans for flexible querying of conversations, agents, and analytics. Use a single endpoint and request only the fields you need. Schema and playground are available in the dashboard for Pro users.

Webhooks Reference

Webhook payloads include an event type (e.g. conversation.created, message.received), a timestamp, and a data object with the resource. We send a signature header so you can verify the request. Retries follow exponential backoff.

Rate Limits

Rate limits depend on your plan: Free/Hobby tiers have lower limits (e.g. 100 req/min), Standard and Pro higher (e.g. 1,000 req/min). Responses include X-RateLimit-* headers. When exceeded, you get 429; back off and retry. Enterprise plans can request custom limits.

SDKs & Libraries

Official SDKs are available for JavaScript/TypeScript (npm install @solveoai/sdk), Python (pip install solveoai), and other languages. They handle auth, retries, and type-safe wrappers for the REST API. See the API Reference for install and usage.

Smart Escalation Rules

Define when to hand off to a human: by keyword, intent, low confidence, or after N messages. In the agent settings, add escalation rules and choose the action (e.g. show “Talk to agent” or create a ticket). Escalated conversations appear in your inbox so your team can take over.

Sentiment Analysis

We analyze sentiment on incoming messages and tag conversations (e.g. positive, neutral, negative). Use this in analytics and in escalation rules—for example, escalate negative sentiment immediately or notify the team when sentiment drops.

Custom Fine-Tuning

Pro+ plans can fine-tune select models on your conversation data for better domain-specific responses. Submit a dataset through the dashboard; we train and deploy a custom model variant for your account. Best for high-volume, specialized use cases.

Workflow Automation

Build multi-step workflows that combine your agent with actions: send emails, create tickets, update CRM, or call external APIs. Use the workflow builder in the dashboard to define triggers (e.g. “when conversation ends”) and steps. Available on higher tiers.

Analytics & Reporting

The analytics dashboard shows conversation volume, response times, resolution rate, and sentiment over time. Filter by agent, channel, or date. Export data for reporting. Use these insights to improve prompts and knowledge base and to tune escalation rules.

Data Privacy

We process conversation and knowledge base data to run your agents. Data is encrypted in transit and at rest. We don’t train our base models on your content. You can delete conversations and knowledge base data from the dashboard or via API; see our Privacy Policy for full details.

HIPAA Compliance

Pro+ and Enterprise plans can enable HIPAA-aligned handling of health data. This includes BAA, stricter access controls, and audit logging. Contact sales to enable and configure HIPAA mode for your organization.

SOC 2 Certification

Our infrastructure and processes are designed to meet SOC 2 Type II requirements. We undergo regular audits and can share certification documentation with Enterprise customers under NDA.

SSO Setup

Enterprise plans support SAML 2.0 and OAuth-based SSO so your team can sign in with your identity provider. Configure the IdP in the dashboard and map roles to SolveoAI permissions. SSO is available for team and workspace sign-in.

Data Residency

You can choose the region where your data is stored (e.g. US, EU). This is configurable at signup or in Settings for Enterprise. Processing stays within the selected region where supported by our providers.