Semantic context layer

Your AI agent doesn't understand your business. Sigil fixes that.

accuracy: 21.9% => 99.5%

Define your metrics, aliases, and exclusions once. Every AI agent query matches your data team's answers. Self-serve. HubSpot-first.

Get started
The gap
Raw CRM access is not understanding
Your AI agent can read every HubSpot field. It still doesn't know what "qualified pipeline" means in your business.
Without Sigil
> "What's our qualified pipeline this quarter?"
$2.4M
# Counted test deals. Used wrong stage. Wrong close date field.
21.9%
accuracy on revenue questions
With Sigil
> "What's our qualified pipeline this quarter?"
$1.87M
# Excluded test deals. Used hs_acv. Correct stage mapping.
99.5%
accuracy on revenue questions
How it works
Three steps. Minutes, not weeks.
No consultants. No week-long onboarding. Connect, define, query.
01

Install the MCP server

Add Sigil alongside your existing HubSpot MCP. Your agent now has two tools: raw data access and business context.

02

Define your context

Your AI agent interviews you about your business: what win rate means, which deals to exclude, how you define pipeline stages. Sigil stores it.

03

Query with accuracy

Every question your agent answers now passes through your definitions. The right filters. The right properties. The right exclusions.

The context model
Everything your AI agent needs to know
Sigil stores four types of business context. Each one prevents a class of errors.
{ metrics }

Metric definitions

What "win rate" or "ACV" actually means in your business. Which fields to use, how to calculate, what to exclude.

metric: win_rate
formula: closed_won / (closed_won + closed_lost)
exclude: partner_deals, renewals
{ aliases }

Field aliases

Your team says "ACV" but HubSpot calls it hs_acv. Sigil maps the language gap so your agent uses the right property.

alias: "ACV"
maps_to: hs_acv
note: "not amount, not arr"
{ exclusions }

Exclusion rules

Test deals, internal accounts, churned-then-returned. The records that silently corrupt every report.

exclude_when: pipeline = "Test"
exclude_when: owner = "Integration User"
applies_to: all revenue queries
{ gotchas }

Known gotchas

The tribal knowledge that lives in someone's head. "Always use fiscal quarters." "Close date was wrong before March."

warning: "close_date unreliable
 before 2024-03-01. Use
 hs_date_entered_closedwon instead."
Architecture
Sits alongside, never replaces
Your AI agent already has a HubSpot MCP for raw data. Sigil adds the context layer that makes those queries accurate.
Your HubSpot MCP

Raw CRM data

Deals, contacts, companies, properties. Everything in HubSpot, exposed to your agent as-is.

Sigil

Your definitions

Metric formulas, field aliases, exclusion rules, known gotchas. The meaning behind the data.

Your AI agent

Accurate answers

Combines raw data with your context. Every query uses the right filters, properties, and calculations.

Result
Data + context = answers that match your data team
Proven in production
Built from a real tool, not a pitch deck
21.9%
99.5%
accuracy on revenue questions

Sigil started as an internal tool at a B2B SaaS company. Deployed company-wide as the primary self-service analytics tool, it answered real revenue questions daily across the entire organization.

The accuracy numbers come from testing real business queries against the data team's known-correct answers, before and after adding the context layer.

Measured across production business queries. Not synthetic benchmarks.

Stop hallucinating revenue numbers

Sign up, define your business context, and your AI agent starts getting it right.

Get started free Log in