Profile Data and Enrichment
AI-to-AI intelligence that gives SI the depth to personalize every interaction
What It Is
Profile data is a distinct layer of contact information that exists alongside standard fields, custom fields, and notes. It lives on the Profile Data tab of the contact detail view.
Unlike contact fields (which have a fixed schema — first_name is always a string, timezone is always an IANA identifier), profile data is freeform. It is stored as JSON key-value records where the keys themselves adapt based on what was being profiled. A profile record from the Data Contextualizer might contain keys like `summary`, `pain_points_identified`, `potential_hooks`, and `deep_desire_questions_frame`. A record from a form analysis might contain `budget_range`, `evaluation_timeline`, and `competitive_alternatives`. A record from a call transcript might contain `sentiment`, `key_concerns`, and `next_steps_discussed`.
Profile data is not designed for human consumption. It is designed for SI consumption. The data structure is optimized for AI comprehension — providing the nuanced, contextual intelligence that SI needs to personalize communication beyond what structured fields can capture.
Why It Matters
Standard fields tell SI *who* someone is. Custom fields tell SI *what category* they fall into. Profile data tells SI *what they actually care about*.
Consider the difference:
- Standard field: `city = Kuala Lumpur`
- Custom field: `primary_job_function = Owner/Founder/CEO`
- Profile data: `pain_points_identified = ["Over-reliance on increasingly expensive Meta ads for lead generation", "Struggling with inconsistent lead flow and low volume (20/month) for the ad spend", "Inefficient lead nurturing and follow-up processes", "Managing a fragmented tech stack is complex and costly"]`
The standard field tells SI where Brandon lives. The custom field tells SI his role. The profile data tells SI *what keeps him up at night*. When SI writes an email to Brandon, it references his specific pain points — not generic messaging about CRM features.
This is the layer that makes SI’s communication feel personally crafted rather than segment-targeted.
How It Works
How Profile Data is Generated
Profile data comes from multiple sources, each producing different key structures:
Data Contextualizer — When raw data is contextualized (during import or via the API), the AI extracts structured intelligence: summary, pain points, potential hooks, reasoning, and deep desire questions.
Form Analysis — When a contact fills out a form, the responses can be analyzed to produce profile data: budget indicators, timeline, current tools, specific interests.
Enrichment Services — External data sources (OSINT, social profiles, company data) produce profile records with company size, technology stack, funding stage, and other business intelligence.
Manual Injection — You can add profile data manually from the Profile Data tab. Enter key-value pairs and a source label (e.g., "Conference Notes", "Discovery Call").
SI Observations — SI can generate profile data based on engagement patterns and behavioral analysis.
Additive, Not Overwriting
Profile data is additive. Each source creates its own record with a timestamp and source label. Multiple profile records accumulate on a contact — a Contextualizer record from July, a form analysis from September, and an enrichment record from October all coexist. SI reads all of them to build a composite understanding.
How SI Uses Profile Data
When SI assembles context for a contact, it reads all profile data records. The pain points inform what SI emphasizes. The potential hooks suggest conversation angles. The reasoning explains why the contact is a fit. The deep desire questions frame the language SI uses.
Because the keys adapt per source and context, SI does not expect a fixed schema. It reads whatever keys exist and incorporates them into its decision-making. A contact with 5 profile records gives SI significantly more personalization depth than one with none.
