AI Automation Features

Everything you need to automate with AI — transparently and under your control. Recipes, workflows, DAG orchestration, prompt studio, knowledge bases, brand voice, scheduling, approvals, AI Bakeoffs, quality monitoring, batch execution, version control, real-time collaboration, and multi-provider LLM access — all in one platform with full audit trails.

Explore our new Platform pages for the complete story

Go to Platform Overview →

Recipes

Reusable AI operations with structured I/O

Recipes are the building blocks of JieGou. Each recipe is a prompt template with defined input and output schemas, so you get consistent, structured results every time.

JSON Schema-based inputs and outputs
Template variables with {{placeholder}} syntax
AI-assisted recipe generation from plain English
Attach documents for additional context
Version tracking and feedback ratings
Community sharing and template library
Image, file, and audio upload with provider-aware handling
Natural-language recipe generation from plain English descriptions

Workflows

Multi-step automation with eleven step types

Workflows chain recipes into sophisticated automations. Sequential or DAG execution, conditional branching, iteration, parallel execution, human approval gates, LLM prompts, eval quality gates, routers, and aggregators give you full control over complex processes.

Recipe steps with automatic input mapping
Condition steps with 8 comparison operators
Loop steps for batch processing
Parallel steps for concurrent execution
Approval steps with email notifications
LLM, Eval, Router, and Aggregator steps for multi-agent orchestration
DAG execution mode with convergence loops and crash-recovery checkpointing
Natural-language workflow generation — describe what you need and JieGou builds the steps
Visual drag-and-drop canvas with zoom, pan, minimap, and snap-to-grid

Scheduling & Triggers

Run automations on your terms

Schedule recipes and workflows with cron expressions, fire them from webhooks, or react to events in real time. Four types of event-driven triggers let you chain runs, monitor emails and Slack, detect connector data changes, and respond to browser events.

Cron-based scheduling with timezone support
Webhook triggers with token authentication
Event-driven triggers — run completion chaining, email/Slack sources, connector data changes, browser events
Google Sheets integration for dynamic inputs
API-based input resolution with JSONPath
Email notifications on success or failure
Output webhooks to push results externally

Conversational AI

Inline AI chat — available on every page

No navigation needed. An inline chat surface is available on every page of the console. Have multi-turn conversations where the AI remembers context, accesses your connected tools, and suggests relevant recipes.

Inline chat surface on every page — no navigation required
Department-aware system prompts
Recipe suggestion chips in conversation
MCP tool access during chat
Persistent conversation history
Document attachment for context
Model selection per conversation

Department Packs

Pre-built automation for every team

Get started instantly with curated packs for Sales, Marketing, Support, HR, Finance, Operations, Legal, Engineering, and Executive teams.

20 department-specific packs
7-10 recipes per pack
2-5 workflows per pack
Suggested schedules included
Integration recommendations
Step-by-step setup guides

Analytics & Observability

Measure what matters

Track every execution across recipes, workflows, and schedules. Monitor costs, success rates, and token usage with department-filtered dashboards.

Execution metrics and success rates
Token usage by provider and model
Cost estimation and overage tracking
Department-filtered analytics views
Template health scoring
Time-saved ROI estimation

Knowledge & Documents

Give your AI real company knowledge

Upload files, crawl entire websites, or connect 12 external knowledge sources. Built-in Firestore vector search with Redis caching delivers sub-second retrieval. JieGou automatically chunks, embeds, and retrieves the most relevant context for every AI operation. Agent Workspaces add governed persistent memory across workflow executions.

Upload files and import URLs
Website crawl pipeline — sitemap discovery, smart filtering, JS rendering, incremental refresh
Connect 12 external sources — Coveo, Elasticsearch, Algolia, Glean, Pinecone, Vectara, Guru, Confluence, Notion, Zendesk, Google Drive, OneDrive
Built-in Firestore vector search — no external vector DB required
Hybrid retrieval with Redis caching for sub-second warm queries
Automatic chunking and embedding
RAG retrieval for recipes and workflows
Knowledge base grouping with auto-context
Encrypted credentials, health checks, and audit trails for every knowledge source
Scoped injection by recipe, workflow, or department
Freshness tracking and re-indexing
Agent Workspaces: governed cross-workflow persistent memory with provenance tracking

Multi-Provider LLM

Any model, any provider — your choice

Run Claude, GPT-5, Gemini, or open-source models like Llama 4 and DeepSeek from a single platform. Bring your own API keys, connect local inference servers, or use platform-provided keys. Pick a different model for every recipe, workflow step, and conversation.

Anthropic (Claude Sonnet/Haiku/Opus), OpenAI (GPT-5.x, o3/o4-mini), and Google (Gemini 3/2.5) with BYOK
OpenAI-compatible endpoints — connect any custom base URL for fine-tuned or self-hosted models
Certified open-source registry: Llama 4 Maverick, DeepSeek V3.2, Qwen 3 235B, Mistral 3 Large on vLLM
Auto-discovery of local Ollama and vLLM servers — models appear automatically
Per-step model selection in workflows — use the best model for each task
Model recommendation engine scoring success rate, cost efficiency, and speed from your execution history
AES-256-GCM envelope encryption for API keys with per-account HKDF key derivation
Per-provider circuit breakers, priority-based concurrency limits, and unified cost tracking

Brand Voice & Governance

Consistent AI outputs across your organization

Define your brand voice once and apply it everywhere. Set tone, formality, glossary terms, and prompt rules that every AI output respects — with full audit logging and approval policies.

Tone and formality settings
Glossary with preferred and prohibited terms
Reusable prompt fragments
Audit log tracking 30+ action types
Multi-approver policies with escalation
RBAC with department scoping

Real-Time Collaboration

Work together on AI workflows in real time

JieGou goes beyond sharing outputs. Your team can co-browse the console, chat in the context of specific runs, and screen-share to review AI results together.

Presence indicators — see who's online and where they are
Contextual chat on runs, recipes, and workflows
Platform-wide team chat with search
Screen sharing with one-click invite
Follow mode — watch a teammate's cursor in real time
Session recording and replay for async review

AI Evaluation & Bakeoffs

Don't just run AI — measure it

Compare recipes, models, and full workflows side by side. Use LLM-as-judge scoring, multi-judge consensus, and live A/B routing to find the best configuration for every use case.

Recipe vs. recipe and model vs. model comparison
Multi-judge evaluation with inter-rater correlation
Statistical confidence intervals (95% CI)
Live A/B test routing with auto-stop conditions
Synthetic input generation from schemas
Workflow-level head-to-head comparison

Workflow Orchestration

Orchestrate multi-stage automations with DAG workflows

Build directed acyclic graph (DAG) workflows where steps run concurrently based on dependencies. Use SubWorkflowStep to invoke other workflows as steps, giving you composable multi-stage orchestration from a single canvas.

DAG execution mode with visual canvas editor
SubWorkflowStep to compose workflows within workflows
Automatic dependency resolution and concurrent wave execution
Convergence loops for iterative refinement with quality gates
Pattern templates: Critic/Refiner, Specialist Router, Debate/Consensus, Plan/Execute/Verify
Crash-recovery checkpointing with resume from any point

Prompt Engineering Studio

Craft, test, and optimize prompts systematically

Go beyond trial-and-error prompt writing. The Prompt Engineering Studio gives you version tracking, live token budget visualization, a variable inspector, few-shot example curation, and an AI-powered optimizer that suggests improvements.

Full version history with diff comparison
Live token count and budget visualization per model
Variable inspector — see all placeholders and their values
Few-shot example curation with drag-and-drop ordering
AI-powered prompt optimizer with improvement suggestions
Side-by-side output preview across prompt versions

Quality Guard

Continuous quality monitoring for AI outputs

Attach a Quality Guard to any recipe and get automatic LLM-as-judge evaluation on every production run. Set quality thresholds, track trends over time, and get alerts when output quality drifts below your standards.

Automatic LLM-as-judge scoring on every run
Configurable quality criteria and thresholds
Quality trend charts with drift detection
Alerts when scores drop below your baseline
Per-recipe quality dashboard on the detail page
Pairs with bakeoffs for a full evaluation workflow

Batch Execution

Process entire datasets in one click

Upload a CSV or paste a data table, then run any recipe across every row. Track progress in real time, filter results, and export outputs as CSV or JSON for downstream use.

CSV upload or paste-in data table
Run any recipe across all rows in parallel
Real-time progress tracking per row
Filter and search results by status or content
Export results as CSV or JSON
Configurable concurrency and retry settings

Workflow Version Control

Ship changes safely with version control and canary rollouts

Every workflow edit creates an immutable version. Compare any two versions with a visual diff engine. When you're ready to deploy, use canary rollouts to send a percentage of traffic to the new version before promoting it to production.

Immutable version snapshots on every save
Visual diff engine for step-by-step comparison
Canary rollouts — route a percentage of traffic to the new version
One-click promotion from canary to production
Rollback to any previous version instantly
Version history with author and timestamp metadata

Instagram DM Automation

Automate Instagram DMs with AI

Connect your Instagram Business account to JieGou and let AI handle incoming DMs. Triage messages by priority, auto-respond to FAQs with confidence scoring, draft contextual replies, and escalate complex conversations — all with full governance controls.

Instagram Graph API webhook integration with HMAC-SHA256 signature verification via Meta platform
FAQ knowledge base with automatic chunking, embedding, and RAG retrieval
Confidence-scored auto-replies — high-confidence answers sent instantly, low-confidence escalated
Story reply handling — process replies to your Stories as engagement or support triggers
Media attachment support — images, videos, audio, and stickers in both directions
Pre-built Instagram Support Pack with triage, FAQ bot, response drafter, and escalation recipes
Multi-language support — AI adapts response language to match the incoming message
Full audit trails, approval gates, and governance controls

Persistent Memory

AI agents that remember — 5 layers of persistent memory

Give your agents institutional memory that compounds over time. A 5-layer hierarchy mirrors how organizations store and retrieve knowledge — from individual entity facts to cross-workflow insights.

Entity Memory — persistent facts about customers, products, and projects with LLM compaction
Workflow Memory — per-workflow accumulated knowledge from execution history
Department Memory — CLAUDE.md-equivalent institutional knowledge per department
Agent Memory — per-agent persistent state across conversations and tasks
Cross-Workflow Memory — entity memory shared across all workflows in a department
LLM compaction prevents unbounded growth while preserving signal
Full governance integration — audit trails, RBAC, and department scoping

Department Memory

Institutional knowledge for every AI agent

Department Memory gives every agent institutional context about its department. Auto-populated from installed recipes, templates, and active workflows. Always-on — not query-triggered like a knowledge base.

Auto-populates from installed recipes, templates, and active workflows
Always-on context injection into every agent interaction and workflow execution
Per-department scoping — Marketing, Support, Finance, Sales, Engineering each get their own memory
Grows as new workflows are added and agents learn from execution
Not a knowledge base — structured context (rules, preferences, procedures), not document retrieval
The CLAUDE.md equivalent for enterprise departments

Facebook Messenger Automation

Automate Facebook Messenger with AI

Connect your Facebook Page to JieGou and let AI handle Messenger conversations. Triage messages, auto-respond to FAQs, generate Quick Reply buttons for guided interactions, and manage Persistent Menus — with full governance and audit trails.

Messenger Platform API webhook integration with HMAC-SHA256 signature verification via Meta platform
FAQ knowledge base with automatic chunking, embedding, and RAG retrieval
Confidence-scored auto-replies — high-confidence answers sent instantly, low-confidence escalated
Quick Reply buttons — AI generates guided conversation options instead of open-ended text
Sender Actions — typing indicators and read receipts for natural conversation pacing
Persistent Menu management for always-available navigation options
Pre-built Messenger Support Pack with triage, FAQ bot, response drafter, and escalation recipes
Postback payload routing — handle menu selections and button clicks as workflow triggers
Full audit trails, approval gates, and governance controls

LINE Integration

Automate LINE customer support with AI

Connect your LINE Official Account to JieGou and let AI handle incoming messages. Upload FAQ knowledge bases, auto-match customer questions with confidence scoring, and respond with rich Flex Messages — while routing low-confidence queries to human agents.

LINE Messaging API webhook integration with HMAC-SHA256 signature verification
FAQ knowledge base with automatic chunking, embedding, and RAG retrieval
Confidence-scored auto-replies — high-confidence answers sent instantly, low-confidence escalated
Rich Flex Message responses with cards, buttons, images, and carousels
Free reply-token responses within 60-second window, push message fallback
Auto-capture: successful responses feed back into the FAQ knowledge base
Pre-built LINE Support Starter Pack with triage, FAQ bot, response drafter, and escalation recipes
Multi-language support — Japanese, Thai, Traditional Chinese, and English
Full audit trails, approval gates, and governance controls

AI Models

Choose the right model for every task

Access 9 LLM providers with full cost transparency. Switch models per recipe or workflow step. Use AI Bakeoffs to prove which model works best for each workflow. Your API keys, your data — JieGou never stores or trains on your content.

Anthropic

Claude Opus 4.6
Claude Sonnet 4.6
Claude Haiku 4.5

OpenAI

GPT-5.2 / GPT-5.1
o3 / o4-mini
GPT-5-mini / nano

Google

Gemini 3.1 Pro / Flash
Gemini 2.5 Pro / Flash
Gemini 2.5 Flash Lite

Mistral

Groq

xAI (Grok)

AWS Bedrock

Azure OpenAI

OpenAI-compatible

Explore all model options including open-source →

Switch between providers per recipe or workflow step

Workflow Engine

Ten powerful step types

Compose any automation logic with recipe, condition, loop, parallel, approval, LLM, eval, router, aggregator, and coding agent steps.

Recipe Step

Execute an AI recipe with automatic input mapping from previous steps. Includes retry with exponential backoff.

Condition Step

Branch execution based on value comparisons. 8 operators including equals, contains, greater_than, and is_empty.

Loop Step

Iterate over a collection, running nested steps for each item. Automatically aggregates iteration outputs.

Parallel Step

Run multiple branches concurrently. Each branch gets an independent output snapshot for isolation.

Approval Step

Pause for human review with full transparency. Approvers receive email notifications, see the complete execution context, and can provide input before the workflow continues.

LLM Step

Issue a direct LLM prompt without a recipe wrapper. Ideal for lightweight transformations and quick decisions.

Eval Step

Score outputs against quality criteria. Trigger convergence loops for iterative refinement until quality thresholds are met.

Router Step

Classify inputs with an LLM and branch to specialist routes. Each route executes a different downstream path.

Aggregator Step

Combine outputs from parallel branches via merge, best-of, or synthesize strategies.

Coding Agent Step

Execute a sandboxed coding agent that writes, tests, and iterates on code within Docker containers with V8 isolate safety. Ideal for data transformations, report generation, and programmatic tasks.

Deep dives

Explore our newest capabilities

Click through for full details on these flagship features.

See it in action

Start building your first automation in minutes.

Talk to Sales