Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/vrashmanyu605-eng/Agentic_Sales-Markerting/llms.txt

Use this file to discover all available pages before exploring further.

Your Ideal Customer Profile (ICP) is the primary filter that determines which leads receive high scores and continue through the pipeline. The ICP matching agent compares every discovered lead against your ICP definition and produces a structured score with detailed attributes. A well-written ICP directly improves the relevance of outreach emails, proposals, and next-step recommendations downstream.

​
Setting your ICP in main.py

The ICP is set as a plain text string in the ideal_customer_profile field of initial_state in main.py: 
# main.py

initial_state = {

    # ...

    "ideal_customer_profile": """
    Mid-size and enterprise companies investing in AI,
    digital transformation, and operational efficiency.
    """,

    # ...
}
The value is passed directly to the ICP matching agent as-is. There is no schema to conform to — write it as a description of your best-fit customer.

​
What makes a good ICP definition

The more specific your ICP, the more precisely the agent can evaluate each lead. Include as many of these attributes as apply to your business:
  • Company size — target segment by headcount or revenue tier (e.g., 200–2,000 employees, mid-market, or enterprise).
  • Industry or vertical — manufacturing, logistics, fintech, healthcare, retail, or another specific sector. Narrow verticals score more accurately than broad ones.
  • Technology maturity — whether the company is actively investing in AI, cloud infrastructure, or workflow automation. Companies already on this path convert faster.
  • Business goals — digital transformation, operational efficiency, cost reduction, or revenue growth. Align this to the outcomes your services deliver.
  • Budget signals — willingness to spend in a specific range (e.g., 10k–10k–50k per project). This helps the agent evaluate revenue_potential and buying_probability.
  • Geographic region — include a region if your sales team operates in specific markets (e.g., North America, Western Europe). Leave this out if you serve globally.
Be specific in your ICP. A vague profile like “companies that want to grow” gives the agent nothing to evaluate against — and produces low-confidence scores across all leads. The more precise your definition, the more useful the matching_attributes and missing_attributes output fields become.

​
What the ICP matching agent evaluates

The agent reads your ideal_customer_profile, the lead_research output from the previous agent, and live industry benchmark data from a web search. It returns a structured JSON object with these fields:
FieldDescription
icp_match_scoreNumeric score indicating how closely the lead matches your ICP
matching_attributesAttributes from the lead that align with your ICP definition
missing_attributesAttributes your ICP requires that the lead does not demonstrate
revenue_potentialEstimated deal value based on company size and scope
urgency_levelHow urgently the lead likely needs your services
buying_probabilityLikelihood of the lead converting based on ICP fit
best_service_fitWhich of your services is the strongest match for this lead
The full output is stored in the icp_analysis field of state and is available to every subsequent agent in the pipeline.

​
Example ICP definitions

IT services for manufacturing companies: 
"ideal_customer_profile": """
Mid-size manufacturing companies (500–5,000 employees) in North America
investing in factory automation and supply chain digitization.
Looking to reduce manual processes and integrate AI-powered quality control.
Budget range $20k–$100k. Currently using legacy ERP systems and seeking
to modernize their technology stack.
"""
AI automation for financial services: 
"ideal_customer_profile": """
Regional banks, credit unions, and fintech companies with 100–1,000 employees
seeking to automate compliance reporting, fraud detection, and customer
onboarding workflows. Comfortable with cloud-based AI solutions.
Budget range $15k–$60k per engagement. Prioritize data security and
regulatory compliance in any technology investment.
"""
Digital transformation for logistics providers: 
"ideal_customer_profile": """
Mid-market logistics and freight companies (200–3,000 employees) dealing with
manual dispatch, routing inefficiencies, and fragmented visibility across
shipments. Actively exploring AI-powered operations management.
Willing to invest $10k–$50k for workflow automation and real-time tracking
systems. Primarily US and EU markets.
"""

​
Target industry

The target_industry field in initial_state works alongside ideal_customer_profile but serves a different purpose: it feeds directly into the web search query the ICP matching agent runs to fetch industry benchmarks. 
# main.py

"target_industry": "IT Services / AI Automation",
The agent constructs the benchmark search query as: 
Ideal Customer Profile benchmarks for {target_industry} IT services
This means setting target_industry to "manufacturing" causes the agent to fetch ICP benchmarks specific to manufacturing IT buyers, which grounds the scoring in real-world industry data rather than generic criteria. Set target_industry to the vertical your leads operate in — not the vertical your own company is in.

Build docs developers (and LLMs) love

Get started for freeTalk to us