IntakeIQ
Intake Intelligence

The intelligent intake layer for modern law firms.

IntakeIQ pre-qualifies every case with multi-model consensus before you pick up the phone. Eight practice areas. HIPAA-compliant. Fewer dead-ends, sharper docket.

47s
median time to score
AI models per case
up to 94%
fewer dead-end consults
Trusted by
Ghalawanji Law
Customer · 2026
Method

How a form becomes a verdict.

Median 47s, end-to-end.
01 Intake
STEP 4 / 7
Tell us about the accident.
Other driver
I was at fault

A claimant fills your form.

Branching logic adapts to practice area. Eight rubrics, dynamic follow-ups, document upload — all white-labelled to your firm.

Avg. 4 min · 22 fields · embed or hosted
02 Consensus
CONSENSUS · 3 MODELS
CLAUDE 81
GPT-4O 76
GEMINI 77
VERDICT 78 /100

Three models score in parallel.

Claude Sonnet, GPT-4o and Gemini each return a 0–100 score with cited evidence. The engine weights by confidence and flags disagreement above 15 points.

Bedrock · Azure · Vertex · zero retention
03 Verdict
IIQ-SAMPLE · DOSSIER
78 Strong — recommend
Hospital records same-day
Police report on file
Commercial $2M policy

Your attorney sees a clean dossier.

Strengths, watch-items, statute checks and Clio sync — landed in the right intake queue before the prospect refreshes the thank-you page.

Clio · Cal.com · Slack · webhook
Documents · HIPAA pipeline

Drop the medical record. We handle the rest.

Pre-signed S3, encrypted at rest, classified at intake, audit-trailed end-to-end. PHI never touches the app server.

DROP FILES HERE · PDF · JPG · HEIC
ER.pdf
→ Pipeline · live
HIPAA
Pre-signed upload S3 · us-east-1
Encrypted at rest AES-256 · KMS
PHI classified medical_records bucket
OCR + extraction Textract · 2.4 MB · 8 pp.
Audit logged CloudTrail · evt #4421
Receipts

Every score, with its receipts.

Three frontier models read the file in parallel. Their answers, their evidence, and their disagreements — exposed, not buried.

Claude · GPT-4o · Gemini
weighted by confidence
variance threshold σ 15
I · Variance & agreement
σ 6.1
low — well within bounds
OK
σ 0 median 6.4 σ 30 · flag

When the three models disagree by more than σ 15, the case is flagged for an attorney — never auto-decided.

II · Citations · 12 supported, 0 unsupported
IIQ-SAMPLE
01
Commercial defendant carrying high coverage.
universal.defendantType = "commercial" · policy.umbrella = $2M
strong
02
Emergency-room treatment same-day as incident.
injury.firstTreatment = 2026-05-03 · incident.date = 2026-05-03
strong
03
Statute of limitations clean — 22 months remaining.
jurisdiction.SOL = 24mo · incident.date = 2026-05-03
strong
04
Witness list pending; one bystander statement to verify.
witness.count = 1 · witness.verified = false
moderate

Every claim cites a specific form field. Unsupported claims are flagged as potential hallucinations — and the case rerouted.

III · Docket · live
SCORING
IIQ-SAMPLE-A 0:42s
Personal Injury 78
STRONG
IIQ-SAMPLE-B 8m ago
Employment 71
STRONG
IIQ-SAMPLE-C 12m ago
Personal Injury 24
DECLINE
IIQ-SAMPLE-D 19m ago
Employment
σ FLAGGED
IIQ-SAMPLE-E 54m ago
Estate 84
EXCEPTIONAL
From the docket

The everyday moments between the headlines.

How the workflow looks in practice — the calls that didn't happen, the dossiers that landed before sunrise, the case files that handled themselves.

SOC 2 ready
HIPAA · BAA available
Customer story
“I’m really impressed with intake IQ. I was struggling with case intake. I was going into consultations with minimal information and spending too much time filling in the gaps during my consultations. Intake IQ gathered the information I needed beforehand so I could schedule the intakes that were worth it. Intake IQ’s AI powered dynamic questionnaire automatically asks the necessary follow up questions based on the information the lead provides. I’m now free to schedule the consults that make sense and my assistants are available to help with other things. I’ll be using this for my practice moving forward.”
Zanah Ghalawanji
Zanah Ghalawanji · Ghalawanji Law
Customer

Ghalawanji Law

Live on IntakeIQ since 2026.
Person on a phone call outside, dictating intake details.
Scenario 01
Friday · 7:42 pm

A plaintiff dictates from a parking lot.

IntakeIQ recognises a strong MVA pattern, runs three-model consensus, and pings the on-call partner before she pulls out of the lot.

47s median time to score
Attorney reviewing case files at a desk in the morning.
Scenario 02
Tuesday · 9:15 am

Five overnight dossiers, ten minutes.

A managing partner reviews everything that scored above 65 before her first call. Three accept, two decline, none waste her morning.

AI models per case
Stacked case files representing PHI handled through the HIPAA pipeline.
Scenario 03
Wednesday · 4:00 pm

The medical record never touched the app server.

Pre-signed S3 upload, AES-256 at rest, PHI classified at intake, audit-trailed end-to-end. The encryption finished before the call did.

HIPAA pipeline · audited
Ready

Sharper docket. Fewer dead-ends.

Twenty minutes with a founder. We'll plug IntakeIQ into your last quarter's intake and tell you what would have been declined.