AgentSDR Review (2026): Open-Source AI SDR for Email, LinkedIn, and WhatsApp

Open-source AI SDR for email, LinkedIn, and WhatsApp: outbound sequences, AI CRM, unified inbox, and lead enrichment — MIT-licensed, self-host or use the hosted service

Category: marketing · Pricing model: freemium (hosted pricing unverified) · Official site: agentsdr.ai

What AgentSDR does

AgentSDR is an open-source (MIT-licensed) AI SDR — an autonomous sales development rep — for outbound across email, LinkedIn, and WhatsApp. Channels: email sequences (cold email from your own Google Workspace mailboxes), LinkedIn automation (invites and follow-ups, safely paced across many accounts), and WhatsApp (recorded calls and message campaigns from your own number). On top: an AI CRM (every reply is classified and the lead is moved for you), a unified inbox (email, LinkedIn, and WhatsApp replies with drafts ready), and data & insights (a lead database shared by every channel, plus enrichment tables with AI columns, APIs, formulas, and Apollo in a grid). The pitch is “replace five subscriptions with one open-source workspace”: run the whole stack on your own server with your own AI key (Docker, Postgres, your env), or use the hosted service. It targets founders (founder-led outbound without an SDR team), agencies (one deployment, a separate organization per client), and sales teams (shared pipeline, per-rep accounts, one inbox). The code lives at github.com/Kandid-ai/AgentSDR.

Who it’s for

Founder-led B2B teams doing outbound without hiring SDRs; agencies running outbound for many clients from one deployment; sales teams that want a multi-channel pipeline with AI reply handling.

Key features

• Email sequences from your own Google Workspace mailboxes
• LinkedIn automation (invites and follow-ups, safely paced)
• WhatsApp campaigns and recorded calls from your own number
• AI CRM: replies classified and leads moved automatically
• Unified inbox with drafts ready across all three channels
• Lead database and enrichment tables (AI columns, APIs, formulas, Apollo)
• Per-channel analytics (replies, meetings, customers)
• Open source under MIT: self-host with your own AI key (Docker/Postgres) or use the hosted service
• Multi-organization: separate org per client (agencies), per-rep accounts (teams)

How to use AgentSDR

1. Self-host from GitHub (Kandid-ai/AgentSDR) with your own AI key, or sign up for the hosted service.
2. Connect your Google Workspace mailboxes, LinkedIn, and WhatsApp number.
3. Build sequences and define your outreach (personas, channels, pacing).
4. Let the AI classify replies and draft responses; approve them in the unified inbox.
5. Track replies, meetings, and customers per channel.

Limitations

• Self-hosting is your ops burden (Docker, Postgres), and your own AI key means your model costs.
• Hosted plan pricing was not published at check time (no pricing page) — treat as unverified.
• Cold email and LinkedIn automation are platform-rate-limited; pacing is built in, but deliverability is on you.
• Young project (a small repository at check time) — expect breaking changes.
• WhatsApp business rules and carrier constraints apply.

Pricing

Could not be verified on 2026-10-09: the code is open source under the MIT license (self-hosting is free, with your own AI key), but agentsdr.ai had no published pricing page for the hosted service at check time. Confirm hosted plan pricing on the site before comparing.

Best for

Founder-led B2B teams that want multi-channel outbound (email + LinkedIn + WhatsApp) with an AI CRM, owned and self-hostable, instead of five separate SaaS subscriptions.

Alternatives

• Zapier — automation glue across your existing tools; no built-in SDR.
• Copy.ai — marketing content generation; no outbound sequences.
• Gemini Notebook — a research assistant, a different stage of the funnel.