Kaaj AI
Also known as: Kaaj Technologies
Agentic AI underwriting OS for small business lenders that turns messy borrower packages into decision ready, policy grounded credit analysis in minutes.
Kaaj is an underwriting platform for small business lenders and brokers that turns borrower packages into decision ready credit analysis. Deals arrive by forwarded email, an embedded application form, portals, cloud drives or an API, and an Inbox Agent triages them. Named agents then take each deal through the same steps.
A Parser Agent extracts fields from documents, statements and emails, a Diligence Agent checks state filings, web presence, addresses and public records, a Credit Agent analyzes cash flow, debt, anomalies and collateral against the lender's policy, a Memo Agent drafts a source backed credit memo with findings and exceptions, and a Pipeline Agent tracks the deal to a memo ready state.
Lenders configure their own credit policies, document requirements, risk thresholds, exception rules and memo formats. Underwriters keep the decision. Exceptions and risk factors are surfaced for them, and their edits and overrides are kept alongside every extraction's evidence in an activity log that reviewers and examiners can follow. Results sync into Salesforce, Microsoft Dynamics, HubSpot, Leasepath, LTi Aspire, Solifi, nCino and other loan origination systems.
Kaaj states SOC 2 Type II compliance, runs on Microsoft Azure and prices by a platform fee with included application volume and usage overages instead of per seat.
Vendor details
Canonical URL
https://kaaj.ai
Category
Enterprise operations agent
Funding status
Seed round of 3.8 million dollars (November 2025) led by Kindred Ventures, with Better Tomorrow Ventures, Karman Ventures, Pythia Ventures, Coughdrop Capital, and others, per Business Wire and TechCrunch. Founded 2024 in San Francisco by Utsav Shah (CEO) and Devangana Sharma, both with backgrounds in credit, fraud, and AI systems at firms including American Express and Varo Bank. The platform has processed over 5 billion dollars in loan applications.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Ingests deals via forwarded email, an embedded application form, portals, Google Drive, Dropbox or an API, parses structured and unstructured documents, and returns results into existing CRM and loan origination systems such as Salesforce, Microsoft Dynamics, HubSpot, Leasepath, LTi Aspire, Solifi, nCino and Jack Henry, with status webhooks.
In practice
Your underwriters spend hours keying in borrower packages that arrive by email. Kaaj's Inbox Agent picks up each forwarded package, and the Parser, Diligence and Credit agents take it through verification and analysis against your configured credit policy.
An examiner asks how a credit decision was reached. Kaaj's activity log shows each deal's steps, and its extraction results, analyst edits, overrides and evidence links trace back to the source documents.
Your team works deals in Salesforce or nCino. Kaaj pushes structured fields, documents and the credit memo into the loan origination system, and webhooks report status changes.
Sources & related URLs
Research sources
Agentic Index coverage score
7.0 / 14 capabilities · 50%
| Integrations & Tool Calling | Full |
|---|---|
|
Structured fields, documents and memo outputs go into Salesforce, Microsoft Dynamics, HubSpot, Leasepath, LTi Aspire, Solifi, nCino, Jack Henry and other loan origination workflows. Packages come in from email, portals, Google Drive, Dropbox and an API. Sourcekaaj.ai/productread 2026-09-28 |
|
| Workflow Orchestration | Full |
|
Six named agents (Inbox, Parser, Diligence, Credit, Memo and Pipeline) hand each deal through intake, verification, analysis, review, memo and sync. The product sets the order of these steps. Sourcekaaj.airead 2026-09-28 |
|
| Knowledge Grounding & RAG | Full |
|
Each memo section is grounded in the deal's source documents, extracted data, business verification signals from state filings, web presence and public records, and the lender's configured credit policy, with evidence links preserved for every extraction. Sourcekaaj.airead 2026-09-28 |
|
| Human Oversight & Guardrails | Partial |
|
Credit decisions stay with the lender's underwriters. Policy exceptions and risk factors are surfaced for analyst attention, and analysts can edit and override outputs. Kaaj names no step where a person must approve before results sync into the CRM or loan system. Sourcekaaj.ai/faqsread 2026-09-28 |
|
| Security, Identity & Governance | Partial |
|
The platform is SOC 2 Type II compliant by Kaaj's account, and its trust center at kaaj.trust.site covers SOC 2 and Microsoft Azure hosting. Kaaj names no SSO, roles or permissions model for the customer's users. Sourcekaaj.ai/data-securityread 2026-09-28 |
|
| Observability & Auditability | Full |
|
A live activity log shows each deal moving through ingestion, parsing, business verification, bank analysis, credit review and memo drafting, and Kaaj keeps extraction results, analyst edits, overrides and evidence links with document level traceability that reviewers and examiners can follow. Sourcekaaj.ai/productread 2026-09-28 |
|
| Memory & State Persistence | Not documented |
|
Deal rooms hold each deal's records. Kaaj names no memory that the agents keep from one deal to the next. Sourcekaaj.ai/faqsread 2026-09-28 |
|
| Deployment & Data Residency | Not documented |
|
The service is hosted on Microsoft Azure. No hosting region is published, and Kaaj names no region choice, customer cloud or on premises option. Sourcekaaj.trust.siteread 2026-09-28 |
|
| Prebuilt Agents / Templates / Packs | Partial |
|
Named agents and modules cover document intelligence, business verification, bank statement analysis, credit memo automation and fraud detection, with segment solutions for equipment finance, brokers, merchant cash advance and community banks. The agents are stages of one underwriting pipeline, not products that each stand alone. Sourcekaaj.airead 2026-09-28 |
|
| Triggers & Channel Coverage | Full |
|
An Inbox Agent watches inboxes, portals, forms and APIs and triages incoming deals. A package forwarded to the lender's Kaaj address, submitted through an embedded form or posted to the API starts the workflow as soon as it arrives. Webhooks report status changes. Sourcekaaj.airead 2026-09-28 |
|
| Model Flexibility & Routing | Not documented |
|
The models Kaaj uses are not named, and its trust center subprocessor list carries no model provider. Kaaj names no model choice for customers. Sourcekaaj.trust.siteread 2026-09-28 |
|
| APIs / SDKs / MCP Extensibility | Partial |
|
A REST endpoint (POST /v1/applications) comes with webhook subscriptions, and Kaaj supports API and webhook workflows. No API reference, authentication scheme or access path to one is published. Sourcekaaj.airead 2026-09-28 |
|
| Testing, Debugging & Optimization | Not documented |
|
Analyst edits and overrides are recorded. Kaaj names no way for a lender to test or score the agents' extractions, analyses or memos, and the regression testing it runs guards its own software. Sourcekaaj.ai/productread 2026-09-28 |
|
| Browser / Computer-use | Not documented |
|
Documents are parsed, public records checked and results synced through integrations. Kaaj names no browser, desktop or computer control. Sourcekaaj.ai/productread 2026-09-28 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Pricing
Custom (contact sales); platform fee plus included application volume
platform fee with included application volume, usage overages, optional implementation fees; not per seat
What is public
The pricing structure (platform fee, included volume, overages, implementation fees, four packaging models) is public; no figures.
Cost watchouts
Overages scale with applications processed or documents analyzed, and implementation and integration fees are separate.
Variable cost rationale
Likely scales with application volume or deals processed given the platform's value proposition of handling far more applications with the same staff; unit not retrieved this session.
Sales call required
Yes, required for paid access
Free / trial
No free trial; workflow pilot by arrangement
Key ambiguities
No figures for the platform fee, included volume or overage rate are published.
Related vendors
- 11th Estate — Agentic platform whose AI engine scans markets, matches an…
- Abacum — FP&A platform for growing companies whose AI Modeler builds, edits…
- Adopt AI — AI native accounting and bookkeeping firm whose agents run…
- Aera Technology — Decision intelligence platform where an always on agent executes…
- Akro AI — On premise operational intelligence that automates document heavy…
- alfred_ — AI executive assistant that triages the inbox overnight and scores…
Alternatives to Kaaj AI
The closest documented capability profiles to Kaaj AI among enterprise operations agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Deposco6.0 / 14A lighter documented profile than Kaaj AI
- GoComet7.0 / 14Fuller documented coverage on Prebuilt Agents, Templates & Packs
- hireEZ8.0 / 14Adds documented Testing, Debugging & Optimization
- Suplari8.0 / 14Fuller documented coverage on Human Oversight & Guardrails and Prebuilt Agents, Templates & Packs
- Basis7.5 / 14Fuller documented coverage on Human Oversight & Guardrails and Security, Identity & Governance
- Byron8.5 / 14Adds documented Memory & State Persistence
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded