ASagentic-sales
00/core / autonomous sales
◆ eval-gated ≥ 80%open source

Autonomous AI agents that discover, qualify, and close B2B deals

Specialized LangGraph agents run your sales motion — discover companies from job postings, enrich profiles with grounded DeepSeek reasoning, find decision-maker contacts, and draft personalized outreach. Every send waits for your approval. Cron-driven, traced, and eval-gated.

discovery agent: sourcing job postingsdiscover → enrich → find contacts → draft → you approve → send → follow up → stop on reply
6
touch cadence
4/7/7/7/7-day gaps, stop-on-reply
100%
human-approved sends
draft-first — nothing sends without you
80%
eval accuracy gate
deepeval bar on every prompt/model change
24x7
autonomous agents
cron-driven langgraph threads
01/pipeline

Seven autonomous stages, one coherent pipeline

from job postings to human-approved outreach — on DeepSeek + Cloudflare D1.

00

Orchestration

[ orchestrate ]

LangGraph Python agents, checkpointed in Cloudflare D1 — resumable, durable, one thread per campaign + contact

01

Discovery

[ discover ]

company_discovery agent sources companies from job postings — grounded, no synthetic leads

02

Enrichment

[ enrich ]

company_enrichment + problems agents reason over each profile with DeepSeek V4-Pro via the Cloudflare AI Gateway

03

Contacts

[ resolve ]

contact_discovery + find_decision_maker agents identify and verify decision-maker emails

04

Scoring + QA

[ score ]

opportunity classifier + company_qa audit score and gate every lead against grounding checks

05

Outreach

[ outreach ]

durable campaign engine drafts a 6-touch sequence — draft-first, sent only on your approval via Cloudflare Email

06

Evaluation

[ evaluate ]

DeepEval ≥ 80% accuracy gate on every prompt/model change, with full LangSmith tracing across agents

02/how it works

The real stack, no fabricated benchmarks

No local GPU farm, no invented numbers — just LangGraph agents, DeepSeek reasoning, and Cloudflare D1, eval-gated and traced.

80%

eval accuracy gate

DeepEval bar enforced on every prompt/model change

100%

human-approved sends

draft-first — nothing sends without your approval

6

touch cadence

4/7/7/7/7-day gaps · stop-on-reply

D1

single source of truth

cloudflare d1 (sqlite) · drizzle orm

RAG

qdrant retrieval

agentic_rag · fastembed bge-small

Trace

langsmith tracing

every agent run traced end-to-end

GraphQL

typed api

apollo server 5 · codegen contracts

OSS

open source

fork it, self-host it, extend the agents

03/core capabilities

Three systems behind the autonomous pipeline

A / 03

Autonomous LangGraph agents

[ checkpointed & resumable ]

Discovery, enrichment, contact-finding, and outreach run as Python LangGraph agents — checkpointed in Cloudflare D1, so every thread is durable, resumable, and cron-driven. No manual babysitting.

one durable thread per campaign + contact, resumed on schedule
  • ~36 specialized graphs registered in the backend
  • interrupt-and-resume keeps long sequences alive across days
B / 03

Grounded, eval-gated AI

[ ≥ 80% accuracy gate ]

DeepSeek V4-Pro reasoning routed through the Cloudflare AI Gateway, with schema-constrained outputs and grounding checks. Every prompt or model change is tested against a DeepEval accuracy bar before it ships.

no prompt/model change ships below the 80% eval gate
  • schema-constrained outputs, validated against the taxonomy
  • full LangSmith tracing across every agent run
C / 03

Human-approved outreach

[ draft-first, always ]

A durable 6-touch campaign engine drafts every email, then waits. Nothing sends without your approval, the sequence stops the moment a reply lands, and delivery runs through Cloudflare Email.

draft-first sends · stop-on-reply · 4/7/7/7/7-day cadence
  • approve, edit, skip, or reject each touch before it sends
  • suppression, caps, and cadence guardrails on every campaign
04/architecture
frontendtyped GraphQL UI on Vercel
Next.js 16React 19Apollo Client
api / dataGraphQL over D1 (SQLite)
Apollo Server 5DrizzleCloudflare D1
agentscheckpointed Python graphs
LangGraphDeepSeek V4-ProAI Gateway
opstracing + eval gate + RAG
LangSmithDeepEvalQdrant

Read the architecture in CLAUDE.md and the backend graph registry.

browse the source

Agentic Sales is fully open source — fork it, self-host it, extend the agents for your ICP.

05/get started

Stop managing pipelines. Let agents do it.

Deploy once, run forever. Your agents discover, enrich, score, and draft outreach around the clock — you approve every send.

autonomous discover → enrich → outreach
every send reviewed by you first
fully traced and eval-gated

▍ not ready yet? get pipeline updates

open sourceself-hostablehuman-approvedeval-gated
ASagentic-sales

autonomous B2B sales

Open-source autonomous B2B sales pipeline. LangGraph agents handle discovery, enrichment, scoring, and human-approved outreach on DeepSeek + Cloudflare D1.

“Built by one person who got tired of paying cloud CRMs $10K/year to own their own leads. This is not a startup. There is no pricing page. It is a pipeline that does one thing well: generate qualified B2B leads, with a human approving every send.”

© 2026 agentic sales