Promote Sign in Sign up

Agents

Flutch ✓

AI sales infrastructure for service businesses

Source rating 5.0 (2) 1.4k views 16 saves Freemium
A/B Testing Acceptance Testing Access Control
More categories (4)
Agent Analytics Agent Governance Agent Operations Agents
Visit website Sign in to save

Free + from $49/mo

Privacy Terms

Verdict

A quick read before you dig into the full page.

Best if
AI sales infrastructure for service businesses
Skip if
Not a replacement for full BI; complex reporting will still live in Looker/PowerBI/SQL
Pick this when
Central place to operate multiple conversational agents across a company

Considering Younet instead? Compare side by side

Overview

Flutch deploys AI infrastructure for service businesses — sales agents that capture inbound across every channel, qualify leads against your rules, follow up automatically, and sync to your CRM.

Our team builds a configured industry pack for your vertical (roofing, HVAC, law, med-spa, home services), launches it on your real leads in 7–10 business days, and stays on as your AI-ops team.

  • Product Analytics: Dau/mau, Retention, Funnels, Conversions
  • Observability (per Agent + Org-level): Latency, Tokens, Error Rate, Failures
  • Agent Configuration Versioning: Multiple Versions, Compare, Promote Best
  • Usage Limits: Per User Group Limits By Messages/tokens/budget
  • Alerts: Notify When Budgets/limits Are Near Threshold
  • Channel Integrations: Whatsapp, Telegram, Slack, Discord, Messenger, Instagram
  • A/b Experiments: Compare Agent Versions On Selected Metrics
  • Pre-release Acceptance Tests: Run Scenario Suites Before Rollout
  • Conversation Audit Log (human-readable): Full Dialog + Step-by-step Processing Trace
  • Copilot: Helps Operate The Platform Without Deep Platform Knowledge
  • Agent Creation + Templates: Built-in Templates For Fast Setup
  • Shareable Chat Ui: Create Agent → Share Link → Others Can Chat
  • Knowledge Management: Articles/files; Reuse Kb Across Agents
  • Sdk For Langgraph: Devs Build The Graph; Sdk Wraps Endpoints + Memory Mgmt
  • Model Gateway: Use Built-in Provider Keys; One Billing Surface For Models
  • User Management & Access Control: Private / Company / Public; Roles/permissions
Flutch website

Access & connections

Where you can use Flutch, and what it plugs into.

Platforms

Web

Integrations

Instagram Telegram WhatsApp Discord Slack MCP

API

Public API available

Browse other tools with an API

Capabilities

What this tool accepts, produces, and is strongest at.

Input

  • API
  • Tabular
  • Text

Output

  • Tabular
  • API
  • Text

Highlights

  • Agents
  • API
  • MCP
  • Open Source

Supported features

  • Agents
  • API
  • MCP
  • Open Source

Pros & cons

Pros

  • Central place to operate multiple conversational agents across a company
  • Works for both engineering and operators: technical metrics plus readable conversation logs
  • Supports safe iteration: versioning, A/B comparisons, and pre-release scenario tests
  • Helps keep usage predictable with budgets/limits and threshold alerts
  • Deploys the same agent to multiple messaging channels without rebuilding integrations each time
  • Lets teams share and reuse knowledge bases across multiple agents
  • Allows quick internal trials via templates and a shareable chat UI
  • Integrates well with LangGraph-based workflows via SDK (endpoints + memory handled for you)
Show 3 more
  • Can connect existing LangChain-based agents (e.g., n8n) for analytics
  • Simplifies model access and billing via a single model gateway
  • Supports access control (private / company / public) and basic role-based permissions

Cons

  • Outcome/ROI analytics depends on clean event tracking and reliable CRM/support/booking data
  • Initial setup requires integrations, event mapping, and agreed “success” definitions
  • Attribution to deals can be uncertain in multi-touch or offline sales processes
  • Not a replacement for full BI; complex reporting will still live in Looker/PowerBI/SQL
  • May be too heavy for small pilots, low-volume bots, or a single internal assistant
  • Some enterprises require BYOK / strict vendor allowlists; a shared model gateway may not fit
  • Coverage of external frameworks is incomplete (no native CrewAI or LlamaIndex yet)
  • Messaging channels can add policy/API constraints that affect what’s possible to automate
Show 2 more
  • Deep customization of workflows or data models may still require engineering work
  • Storing conversation logs can trigger security/compliance reviews and retention requirements

Compare alternatives

Similar tools in the catalog — with a clear reason to open each one.

Why look Typically cheaper

Full comparison

Why look Typically cheaper

Full comparison

from $249/mo

Full comparison

Free + from $99/mo

Full comparison

Why look Typically cheaper

Full comparison

Why look Typically cheaper

Full comparison

Releases 1

Initial release September 13, 2025 Current
Initial release of Flutch.

FAQ 12

What is Flutch?
Artem Ignatev
🛠️ 1
tool
wrote:
Flutch is an analytics and governance layer for conversational AI agents. It helps teams operate multiple agents in one place: track performance, manage versions and rollouts, control usage and budgets, and review conversations with an audit trail.
Who is it for?
Artem Ignatev
🛠️ 1
tool
wrote:
Teams running conversational agents in production: engineering/platform teams, AI teams, and operators (RevOps, Support Ops, CX, product owners) who need visibility and control across agents, channels, and versions.
How is this different from LLM observability tools?
Artem Ignatev
🛠️ 1
tool
wrote:
LLM observability focuses on technical signals (latency, errors, tokens, traces). Flutch is built to run agents as a business system: versioning, rollouts, experiments, audits, limits, and analytics that operators can use.
Do you build agents or only monitor them?
Artem Ignatev
🛠️ 1
tool
wrote:
Both are possible. You can create agents using templates inside Flutch, or connect existing agents built with other stacks (e.g., LangGraph / LangChain-based systems like n8n).
Can I connect my existing agent framework?
Artem Ignatev
🛠️ 1
tool
wrote:
Yes for LangGraph and LangChain-based setups (including n8n). Native support for CrewAI and LlamaIndex is not available yet.
What channels can agents run on?
Artem Ignatev
🛠️ 1
tool
wrote:
Common messaging channels including WhatsApp, Telegram, Slack, Discord, Messenger, Instagram (availability may depend on each platform’s API and policies).
What does “agent versioning” mean in practice?
Artem Ignatev
🛠️ 1
tool
wrote:
Each agent has multiple configurations/versions. You can test versions, compare performance, and control rollouts so changes are tracked and reversible.
How do A/B tests work?
Artem Ignatev
🛠️ 1
tool
wrote:
You can split traffic between agent versions and compare results using selected metrics (usage, conversions/funnel steps, quality scores, etc.), then promote the better-performing version.
What are acceptance tests?
Artem Ignatev
🛠️ 1
tool
wrote:
Scenario-based tests you run before releasing an update. They help catch regressions and risky behavior using a predefined set of conversations and edge cases.
Does Flutch offer on-premise deployment?
Yes, Flutch offers assistance for on-premise deployment in any infrastructure, or provides Helm charts for self-deployment.
What is the conversation audit log?
Artem Ignatev
🛠️ 1
tool
wrote:
A human-readable log of the full dialogue plus a trace of how the agent processed the message and produced the response, designed for non-developers as well.
How do cost controls work?
Artem Ignatev
🛠️ 1
tool
wrote:
You can set limits per user group (messages, tokens, budget) and receive alerts when thresholds are approaching.
Flutch Free + from $49/mo
Visit website