Back to examples
Agentic operationsInternal workflow example

Agentic outbound research pipeline

Turn a research brief into checked candidates, outreach drafts and follow-up tracking with approval before sending.

This example explains a workflow design and its controls. It does not represent verified client results. Tooling and benefits must be validated in each project's context.

Agentic outbound research pipeline editorial cover

Solution example

Agentic outbound research pipeline

Turn a research brief into checked candidates, outreach drafts and follow-up tracking with approval before sending.

The design connects these steps: Brief intake → Missing-info loop → Search creation → Candidate pool → Account validation → Deep research → Company score → Discover contacts → Score contacts → Suppression → Draft queue → Approve/send/follow-up. Rules, exceptions and approvals are defined before external actions are connected.

Workflow map

Three independent workflows in one outbound pipeline

Each workflow has its own trigger. They can run one after another, but they can also be used separately for intake, research, or approved sending.

Workflow 0 - Search Brief IntakeTrigger: email, chat, or manual requestWorkflow 1 - Research and Draft QueueTrigger: new search ready for investigationWorkflow 2 - Approved Outreach And Follow-UpTrigger: human-approved draftReplyComplete0. Receive requestEmail | chat | manualChannel1. Validate briefICP | region | constraintsOpenRouterEnoughinfo?2b. Ask detailsMissing | exampleResend2a. Create searchBrief | filters | goalSupabase3. Confirm startSearch acceptedResendSearch starts Workflow 1Allowed1. Load searchBrief | thresholds | limitsSupabaseBatch ruleThe search defines atarget company count.Each company follows its path.One blocked company does notstop the remaining pool.2. Candidate poolDiscover or load companiesTavily3. Validate accountCustomer | duplicate | ownedPython + CRMAllowed toinvestigate?ContinueNext companyQualified4. Deep researchWeb | signals | sourcesTavily5. Company scoreFit | pain | toolsOpenRouterCompanyqualified?ContinueNext companyContact fit6. Discover contactsRoles | email | LinkedInTavily7. Score contactsPersona | authority | angleOpenRouterContactqualified?ContinueNext contactClear8. Suppression checkEmail | domain | rulesPythonSuppressed?ContinueNext contact9. Draft outreachFirst touch | follow-upsOpenRouter10. Human reviewDraft pending approvalSupabase11. Human decisionApprove | edit | rejectApproval starts Workflow 2Clear1. Approved draftReviewer | versionSupabase2. Fetch contextDraft | company | contactSupabase3. Recheck suppressionBefore any sendPythonStillallowed?EndNo send4. Sync HubSpotCompany | contact | noteHubSpot5. Send outreachApproved message onlyResend6. Log activityEmail | timeline | ownerHubSpot7. Save follow-upSent | reply | next dateSupabaseEnd

Main database: Supabase

The workflow uses Supabase as the source of truth: it stores each record, status, and key event so the team can see what happened, retry failures, and debug without searching every tool.

Tool icons

SupabasePython/FastAPITavilyOpenRouterHubSpotResend

Pipeline rationale

This is not an outbound bot. It is a controlled sales operating system.

The agentic part is not a model sending messages on its own. It is the system breaking an ambiguous commercial job into smaller decisions: ask for context, research, discard, score, draft, wait for approval, send and record what happened.

Workflow 0 exists because human briefs rarely arrive perfect.

The system can ask for market, geography, company size, buying signals, exclusions and goal before spending research effort.

Workflow 1 exists because researching is not the same as sending.

This is where the pool is built, customers/duplicates are blocked, company/contact fit is scored and a review queue is prepared.

Workflow 2 exists because sending is an external action.

It only starts after human approval, rechecks suppression, then syncs HubSpot, sends and stores follow-up state.

Why the gates matter

  • Complete brief before investigation starts.
  • Customers, duplicates and owned accounts blocked before deep research.
  • Company must qualify before contacts are discovered.
  • Each contact must qualify before drafting.
  • Suppression checked before draft and again before send.
  • HubSpot and Resend touched only after human approval.

Add-ons that fit on top

Buyer intent and website-visit signals to prioritize warmer accounts.
Probable-tool inference: CRM, email, analytics, helpdesk, ecommerce or internal stack.
Competitor, hiring, funding/news and leadership-change research.
Separate scoring for company, contact, likely pain, timing and ability to pay.
Internal approval UI for editing messages without touching Supabase.
CSV/Google Sheet export for commercial review or reporting.
Follow-up sequences with pauses, limits, reply detection and stop rules.
Reply classification: interested, not now, wrong person, unsubscribe, bounce.
Weekly digest with searches, blocked accounts, approved drafts and replies.
Quality dashboard: sources used, research cost, approval rate and errors.
Owner assignment by territory, industry or existing account portfolio.
Low-cost testing mode with simulated research before using real web search.

The scenario

Design scenario: Turn a research brief into checked candidates, outreach drafts and follow-up tracking with approval before sending. The starting point is a workflow with scattered information, manual steps and responsibilities that need to be made explicit.

Before implementation, we review available data, responsible owners and decisions that need approval. Understand the workflow first.

The approach

The design connects these steps: Brief intake → Missing-info loop → Search creation → Candidate pool → Account validation → Deep research → Company score → Discover contacts → Score contacts → Suppression → Draft queue → Approve/send/follow-up. Rules, exceptions and approvals are defined before external actions are connected.

AI can help classify, extract or draft; sensitive actions retain human review. Clear controls at each step.

What was used

Tooling was chosen from the process outward, not from a pre-decided technical preference. Each piece needed a clear owner, a stable integration path and a simple way to inspect errors.

In practice, the build combined Supabase, Python/FastAPI, OpenRouter, Tavily, HubSpot, Resend. The tools visible to the team stayed close to their daily work, while integration logic was documented and kept separate from sensitive commercial decisions.

SupabasePython/FastAPIOpenRouterTavilyHubSpotResend

Workflow design

Manual approach and proposed workflow

Manual approach

1Human writes target idea
2Rep searches companies manually
3Existing customers checked late
4Contacts researched one by one
5Fit and authority judged from memory
6Message angle written from scratch
7Approval lives in chat or spreadsheet
8Follow-up tracked separately

Proposed workflow

1Brief intake workflow
2Missing-info loop
3Candidate pool
4Account-status validation
5Deep company research
6Company and contact scoring
7Suppression check
8Scored draft queue
9Approval workflow
10HubSpot/Resend send
11Follow-up state

Visible artifacts

  • Search brief record
  • Candidate company table
  • Account-status validation event
  • Company research summary
  • Contact score rationale
  • Outreach draft queue
  • Approval log
  • HubSpot activity
  • Resend message id
  • Follow-up status

Controls

  • Brief must be complete before search starts
  • Existing customers, owned accounts and duplicates are blocked before deeper research
  • Company must qualify before contacts are discovered
  • Each contact must qualify before drafting
  • Suppression is checked before draft and again before send
  • Human approval required before HubSpot or Resend
  • Every score keeps source evidence
  • Follow-up timing is tracked instead of relying on memory

What this aims to improve

The aim is to reduce manual coordination and make records, exceptions and the next action visible. Any improvement must be validated against the project's own data and operating conditions.

Project scoping establishes a baseline and success criteria so proposed improvements can be tested. Validate the benefit in context.

The workflow in one line

01Brief intake02Missing-info loop03Search creation04Candidate pool05Account validation06Deep research07Company score08Discover contacts09Score contacts10Suppression11Draft queue12Approve/send/follow-up

Tooling and boundaries

The design shows how Supabase, Python/FastAPI, OpenRouter, Tavily, HubSpot, Resend can work together. Final tooling, scope and controls are agreed for each project.

Integrations need appropriate permissions, run logs and a way to review exceptions. The team should be able to understand and operate the workflow.

What this example demonstrates

Design components
  • Brief intake workflow
  • Candidate pool creation
  • Account-status validation
  • Company/contact research
  • Company/contact scoring
  • Draft approval queue
  • HubSpot/Resend send
  • Follow-up tracking
Intended controls
  • Brief must be complete before search starts
  • Existing customers, owned accounts and duplicates are blocked before deeper research
  • Company must qualify before contacts are discovered
  • Each contact must qualify before drafting
  • Suppression is checked before draft and again before send
  • Human approval required before HubSpot or Resend
  • Every score keeps source evidence
  • Follow-up timing is tracked instead of relying on memory

Discuss a similar workflow.

Discuss your workflow

Free initial conversation. No obligation.

Ductio newsletter

Practical ideas for better work.

Get ideas on automation, connected systems and more reliable processes.

Ductio
DuctioSYSTEMS
Custom automation, connected systems and support for teams that need reliable, documented workflows.
Operating focus

CRM operations, reporting workflows, approvals, document handling, and AI-assisted internal tools.

Copyright © 2026 Ductio. All rights reserved.

Discuss your workflow

Free initial conversation. No obligation. We will work through the details together.

Add more details (optional)

We use these details to respond to your inquiry. Please do not include passwords or confidential business data. Privacy notice

You can also email contact@ductiosystems.com.