Customer onboarding
Welcome to AI Growth Engine
We help you find customers, convert them, and grow your business with a guided agentic setup.
Demo mode ready
Step 1 of 4
Choose what you want to achieve.
Customer happy path
Generate Seller Leads
See the journey from selected goal to actual seller lead results, without agent complexity.
Generated0
Contacted0
Interested0
Converted0
Platform Happy Path Library
Choose or create a reusable growth path
Each Happy Path should make one business outcome easy to understand, run, approve, and improve.
New Platform Path
Create a Happy Path for reuse
Agent Pillars
Choose agents for this Happy Path
Configure Selected Path
Confirm the business context
Intake data has changed. Some downstream outputs may need review.
Path brief
What the system will run
Journey Cards
Generate Seller Leads
Current Value
Outcome movement
What is working
Strongest signals
What needs attention
Weak points
AI suggestion
Strategic recommendation
Create more seller microsites for Chennai exporters
This segment has the highest intent and strongest contacted-to-interested movement.
Potential impact: +18% qualified leads in 30 daysNext best actions
Prioritized recommendations
Quick action bar
Move the journey forward
Start here
Click on a lead to see details and contact them.
Great job!
You just started converting your first lead.
Next: follow up 4 more leads.
Top leads to contact
Recommended plan
Growth Plan
Rs. 49,999/month + Rs. 1.5 lakh setup
Best for serious lead generation with microsites, lead extraction, CRM, scoring, and sales conversations.
Growth cockpit
Real-time decisions — strategy, execution, and ROI in one surface.
Your control tower prioritises speed and clarity: approve, execute, and optimise without hunting through screens.
Live loop
Next best revenue action
Hot leads, approvals, and optimisation signals surface here first.
Execution health
Synced from workspace API when online.
- Campaigns —
- Failed runs —
- Pending approvals —
- Blocked packages —
Attention
- High CPC or budget risk — review paid media diagnostics.
- Execution blocked — validate readiness before external push.
- Recommendations never auto-spend; human approval required.
Activity
- Insight Generate Phase 8 insights after ingest.
- Approval Governance queue updates on approve/reject.
- Execute External runs only after ticket approval.
Insight command center
Your role-based command center
Nudges active
Business snapshot
Live growth signals
Phase 5 · Controlled execution
Campaign drafts & workspace usage
Product-scoped drafts only. No autonomous send — approvals required before activation or outbound activity completion.
Execution health loads with API (Phase 9D — read-only).
Phase 9A · Paid media & LinkedIn drafts
Campaign plan drafts
Draft-only layers — no ads launched, no LinkedIn messages sent automatically.
Paid media plans
Recommendation only — no automatic execution.
LinkedIn plans
Draft only — no LinkedIn messages sent.
Execution readiness
Prepared for execution — not yet launched. No external APIs; Phase 9C performs pushes.
Phase 9C–D. External execution requires an approved ticket. LinkedIn messaging is export-only in this phase. Live launch is disabled. Connectors and execution run on the server only. Performance sync and ingest are manual — no continuous polling; insights only (no auto spend or auto outreach).
Phase 8 · Revenue intelligence
Insights & decisions (non-autonomous)
Explainable signals and suggestions only — no execution from this layer.
Insights
Recommendations
WIG performance
Microsite alignment flags
Strategy health
Six business blocks, not 32 agents
Demand engine
Microsites
Lead engine
Extraction to scoring
Native GTM CRM
Smart lead table
| Lead / Company | Source | Mission | Score | Status | Owner | Last Activity | Next Action |
|---|
CRM pipeline
Visual movement by stage
Focused command workspace
Choose one intelligence view at a time
My day
Daily worklist
AI recommendations
Next best actions
Growth insights
What is working
Approval queue
Control layer
Mission progress
Lead-to-conversation loop
Business Report Workspace
Choose an outcome and get a business-ready report.
This workspace is built for founders and business teams. You choose what you want to achieve, confirm a few inputs, and the system creates a usable report without showing the agent machinery.
Founder view
Understand My Market
Generate a market intelligence report using approved company context and your latest inputs.
Step 1
Choose Outcome
Step 2
Confirm Inputs
Step 3
Building your business report...
Using your inputs, approved context, and specialist methodology to create a business-ready report.
Step 4
Business Report
Custom Flow Builder
Advanced agent workflow builder.
Use this when you want manual control over agents, order, inputs, approvals, and technical handoffs. Normal users can stay in the Business Report Workspace.
Foundation API stage runs (POST /api/projects/…/stages/…/run) return execution metadata only — Local vs LLM Gateway, provider id, model, usage id, tokens, latency, and failure reason — with prompts and provider secrets kept on the server.
Draft
Custom Growth Flow
No code. Select a goal, add agent steps, review outputs, then write approved learning into memory.
Strategy spine
Align 23-agent workflow & outputs
Start point
Choose blank, template, or an existing Happy Path
Orchestration canvas
Start → Add steps → Run Autonomous Flow
Start
Select goal or blank canvas
Inputs can come from memory or be typed manually before each agent runs.
Mission selection
What type of leads do you want?
Choose the mission. The system will open the right Happy Path with leads, next actions, and results already prepared.
Recommended
Sellers
Best starting point for Troconet marketplace supply growth.
Pick a mission
Every mission routes to a working Happy Path
Campaign and journey orchestration
ICP segments mapped into campaign Happy Paths
Execution
Contact leads and update the CRM without leaving the flow.
Click a lead, add notes, record objections, and let the system suggest the next best conversation action.
Suggested reply
Offer assisted onboarding
This reduces trust risk for hot seller leads.
Pipeline view
Advanced CRM board
Product-linked campaigns
Campaign drafts · channel plans · lead lifecycle
CLM
Customers · workspace API
Customer detail
Lifecycle action timeline
Phase 7 review queue loads when API is connected.
Lead activity timeline
Workspace-scoped · read-only
Real-time lead extraction
Build, preview, clean, and save approved records into CRM
Extraction preview
Records before CRM save
Dedupe and hygiene
Quality decisions
Extraction history
Recent runs
Omnichannel CRM
Capture, interact, call, record, and transcribe from one customer record
Messaging, social, SMS, and email
WhatsApp BusinessConnected - Main WhatsApp - Lead capture
Email inbox / SMTPReady to configure - Outbound nurture and reply tracking
SIP trunk, PRI, IVR, OBD
SIP trunkReady - Inbound enquiry calls - Routed to Priya
IVRDraft - Menu routing by product, geography, language, and customer stage
Recording and transcription store
Latest transcript preview
Recording will be stored with consent, converted into transcript, summarized, scored, and written back to CRM as conversation memory.
IntentHigh
ObjectionNeeds proof
SentimentPositive
Next actionSend proof + callback
Call recordingsRequired before recording - Tenant encrypted storage
TranscriptsAuto detect - objection, intent, sentiment, urgency, next action
Test a channel into CRM
Use this to prove that WhatsApp, SMS, email, Messenger, Instagram, LinkedIn, and voice sources can create CRM-ready leads before live API credentials are added.
No test run yetRun a plugin test to create a CRM lead and activity.
Omnichannel interaction in one flow
Instagram ad clickLead source and campaign tag captured
Web chat / formName, phone, email, requirement saved to CRM
WhatsApp proof shareTemplate sent, delivery and response logged
Email follow-upCase note and buyer proof asset delivered
SMS reminderCallback reminder sent with consent and opt-out
SIP / IVR callCall recorded, transcribed, scored, and assigned
Field mapping and governance
Control CRM schema, compliance, enrichment, and export rules
Smart leads
Click a lead to continue the sale
New18
Hot12
Due today9
Needs human4
Won6
Manual CRM entry
Add a lead captured from call, event, referral, WhatsApp, or offline source
CRM intelligence
Accuracy, source quality, stage movement, and next actions
Hot score is matching response and qualification signals.
Highest contact-to-response movement this week.
Follow-up proof and timing need improvement.
Missing WhatsApp, designation, and company website fields.
CRM configuration
Configure lifecycle, products, services, and CRM-only users
Awareness to customer stages
Every customer can define their own journey stages, so scoring, outreach, and reports match the way their business sells.
Product / service lines
Leads can be mapped to different product lines so each campaign, bot, and salesperson can use the right offer context.
Seller onboardingGTM service - Primary GTM package for manufacturers and exporters
Buyer discoveryMarketplace service - Lead capture and demand validation workflow
CRM access and platform admin
CRM-only users get lead, activity, and pipeline access. The platform admin keeps full control of agents, memory, billing, exports, and governance.
Admin UserEMP-ADMIN - Platform Admin - All functionality
Priya SharmaEMP-001 - CRM Manager - All CRM records
Agent login and access
Login screen for platform admins, CRM users, and marketing consultants
Agent / user login
Role-based access
Platform AdminAll agents, CRM, billing, memory, exports, governance, and setup.
Marketing ConsultantAssigned accounts, campaigns, CRM actions, agent outputs, and client clarifications.
CRM UserLeads, activities, pipeline, conversations, notes, calls, and follow-ups.
Customer UserApprovals, happy path, agent output visibility, and selected CRM records.
Client-specific CRM screen
Keep mandatory fields fixed and choose editable optional fields per customer
Autonomous campaign execution
Upload leads. The system scores, segments, nurtures, adapts, and hands over sales-ready opportunities.
This is not a recommendation dashboard. It is a governed execution engine that decides what to send, when to send it, where to send it, when to pause, and when to escalate to humans.
Waiting for import
Tenant-safe campaign brain
Revenue Memory, Growth Memory, and Control Memory will update after approval.
Bulk lead ingestion
Upload, map, validate, preview, dedupe, and import
Data cleansing and dynamic scoring
Clean, dedupe, classify, and explain every score
Micro-segmentation engine
Dynamic segments based on fit, intent, behaviour, objections, and channel preference
Psycholinguistic nurture intelligence
Message, channel, timing, escalation, and pause rules per segment
Autonomous execution engine
Next action, channel, timing, message, pause, and escalation decisions
Phase 5 readiness
Omnichannel plan · no auto-send
Activities stay planned or waiting_for_approval until a human approves. Third-party channels are never called from this layer.
Governance and approval
Approval workflow, compliance checks, intrusion validation, and human intervention
Learning and optimisation
Tenant-specific channel, timing, fatigue, segment, and conversion learnings
Microsite workflow
Create a customer-ready microsite from approved strategy.
The system turns memory, ICP, positioning, SEO, copy, lead capture, CRM tags, and follow-up logic into one usable microsite output.
Draft
Find Buyers for Your Factory in Coimbatore
Audience: MSME manufacturers looking for buyer demand.
Microsite portfolio
Plan capacity, drafts, approvals, publishing, and recommendations
Phase 7A · Governed assets
Microsite gallery, versions, and review
Human-approved iterations only. New revisions create versions; nothing is published externally from here.
Microsite detail
—
Preview
Brand profile
Source agent outputs
Versions
Alignment review
Run brand and strategy checks for the selected version.
Comments
Inherited strategy
Already understood from approved memory
Recommended opportunities
Choose a conversion asset
Microsite builder
Strategy, copy, SEO, lead capture, and governance
Objective: capture qualified buyer demand for MSME manufacturers.
Live preview
Find Buyers for Your Factory in Coimbatore
Get qualified buyer inquiries, assisted onboarding, and CRM-ready leads without building a marketing team.
Trust block
Proof-led copy, FAQ objections, CTA tracking, and thank-you flow generated.
Created microsite
Your microsite is ready to use
/microsite-url
Fields, source tags, scoring hints, and follow-up activities are connected.
Lead capture, enrichment, scoring, sales outreach, conversation AI, and analytics are ready to continue.
Lead form and embed center
Create forms, landing pages, chat bot, and voice bot embeds
Lead page builder
Create a landing page and lead form with fields of your choice
Select lead form fields
Mandatory CRM fields stay locked; optional fields can be changed per microsite.
Capture buyer enquiries
Built for ICP
Find buyers for your factory
Proof-led value proposition will appear here.
Solutions
One focused conversion path for one clear buyer segment
Problem copy
Solution copy
Proof and buyer confidence
Clear answers buyers can trust before they speak to sales
Proof copy
Lead capture form
CRM-ready lead capture
Every submission writes source, segment, score, owner, and next action.
Conversation intelligence
AI conversation bots for chat, voice, WhatsApp, email, and video follow-up.
Scripts are generated from ICP, micro-segment, psycholinguistic angle, CRM stage, objections, and guardrails.
Guardrails active
No outside-context conversations
Bots stay inside approved product, offer, compliance, and sales boundaries.
Bot capability layer
Channels and use cases
Script lab
A/B testing and learning loop
Live conversation console
Practice or run a controlled customer conversation
Voice bot session
Start the bot without asking for microphone access. Enable live mic only when you want browser voice capture.
CRM and assignment integration
Conversation data becomes lead intelligence
Waiting for actionSend a message or sync the conversation to create CRM memory.
Guardrails
Bounded conversation control
Growth insights
Simple signals. Clear next action.
No complex analytics dashboard. Just what is working, what is weak, and what to do next.
Next loop
Create microsite for exporters in Chennai
This can start the next Happy Path immediately.
Creative asset engine
Image, video, advertorial, social identity, content, and SEO assets inherit approved strategy.
This is not a blank generator. Assets are created from ICP, positioning, offer, campaign objective, brand tone, proof needs, and lead-capture logic.
Review required
Assets cannot publish until approved
Brand, compliance, assumptions, and performance intent are checked first.
Takeover asset queue
Generation-ready creative work
Asset preview
Creative asset
Advertorial PLUS
Conversion narrative draft
Asset review layer
Governance checklist
Launch control
GTM Pro, measurement setup, deep launch QA, export, and post-launch bridge.
Launch is blocked until critical strategy, tracking, lead capture, sales handoff, and compliance checks are approved or intentionally overridden.
82% ready
3 items need review
Pricing assumption, analytics event map, and conversation guardrails.
Launch readiness QA
Deep validation checklist
GTM Pro
Launch plan
Measurement setup
Analytics events
Final export
Launch handoff
Market Signal & Executive Nudging Agents
Events, headlines, and executive nudges
Convert market signals into campaigns, microsites, sales triggers, and decision-ready nudges.
In-House AI Data Agent Layer
Ask, enrich, validate, and learn from tenant data
Tenant-safe intelligence for CRM, campaigns, microsites, market signals, agent memory, executive nudges, and growth decisions.
Tenant memory
Resume previous work
Saved custom flows, generated outputs, human reviews, tenant recommendations, and signal memory are preserved here so work can continue from where it stopped.
Workflow memory
Saved workflow sessions
Learning memory
Human and tenant learning
Signal memory
Market signals and nudges
Agent control center
Universal agent framework, shared memory, orchestrator, governance, and post-launch growth loops.
Every stage follows the same pattern: load memory, identify gaps, build a context pack, draft output, approve to memory, and recommend next stages.
65 stages verified
Takeover-agent framework
Pre-launch, revenue, creative, growth, hardening, and autonomous mode are mapped.
Agent Work Viewer
Inputs, outputs, and handoffs customers can inspect
Agent library
Agent intelligence depth
Hardening layer
Governance and control
Hardening layer
Governance and control
Autonomous learning
Controlled optimisation loop
Autonomous learning
Controlled optimisation loop
Customer subscription
AI Growth OS pricing
Admin editor
Edit Starter
Usage add-ons
Revenue controls
These settings are local for the prototype. In production they move into the tenant admin database with audit logging.
Phase 8B — governed prompts
Prompt IP Lab
Registry, version history, performance-informed proposals, review queue, experiments, and scorecards. Gold prompts are never auto-overwritten; activation requires human approval and safety gates (Super Admin manual override when needed).
Connect the live API to use Prompt IP Lab.
Your session does not have Prompt IP privileges (Super Admin or Prompt Admin).
Prompt registry
Versions & lifecycle
Prompt text (privileged — not for tenant export)
Improvement signals
Prompt intelligence — reviews & scorecard
Review queue tracks proposal risk and human decisions. Scorecard aggregates tenant-scoped performance signals for the selected version (requires API data).
—
Proposals & experiments
Phase 8A.1 / 8A.2 — catalogue & routing
Model provider registry
Super Admin provider list. No LLM calls from this UI — routing preview below is read-only metadata.
Connect the live API to load model providers.
Platform Super Admin only (provider table).
Platform default
Super Admin configuration (read-only here).
—
Tenant / workspace routing
—
Agent routes & resolution
—
Builder console
Private agent workflow foundation
Phase 8A.4
LLM gateway test
Validates routing and gateway execution only. Approved agent runs use the same routing metadata via POST …/stages/…/run (no prompt text or secrets in the browser).
Connect the live API to use the gateway test panel.
Route preview
—
Last run result
—
Recent usage (this workspace)
—
Controlled execution layer
23-Agent GTM OS Workflow
Uses existing library agents only. One run per approval and per click. Does not replace the full catalogue.
Foundation HTTP stage runs return execution mode (Local / LLM Gateway), provider id, model, usage id, tokens, latency, and failure reason — prompts and provider secrets stay server-side.