Back to vendors
K

Kaaj AI

Also known as: Kaaj Technologies

Visit site
Entry priceCustom (contact sales); platform fee plus included application volumeFull pricing detail

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 rather than 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

Small business loan underwritingBusiness verification and KYBBank statement and cash flow analysisCredit memo generation and fraud detection

Target customers

Small business and equipment finance lendersLoan brokersPrivate credit and community lending teams

Deployment options

SaaS

Integrations

Ingests deals via forwarded email, an embedded application form, or API, parses structured and unstructured documents, and returns results into existing CRM and LOS systems such as Salesforce, Microsoft Dynamics, HubSpot, and Zoho, with status webhooks.

Agentic Index coverage score

7.0 / 14 capabilities · 50%

Integrations & Tool Calling Full

Kaaj pushes structured fields, documents and memo outputs into Salesforce, Microsoft Dynamics, HubSpot, Leasepath, LTi Aspire, Solifi, nCino, Jack Henry and other loan origination workflows, and takes packages 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, a multi agent flow whose order of steps is the product's own.

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, which reserves the decision to people; no step where a person must approve before Kaaj syncs results into the CRM or loan system is documented.

Sourcekaaj.ai/faqsread 2026-09-28

Security, Identity & Governance Partial

SOC 2 Type II compliance is stated on Kaaj's security page, and its trust center at kaaj.trust.site lists SOC 2 with Microsoft Azure hosting, but no SSO, roles or permissions model for the customer's users is documented.

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

No memory the agents keep across deals is described; deal rooms hold each deal's records, which is the application's data. An investor's announcement claims every loan makes the next one sharper, but Kaaj's own pages do not say so.

Sourcekaaj.ai/faqsread 2026-09-28

Deployment & Data Residency Not documented

Kaaj runs as a hosted service on Microsoft Azure per its trust center, with no region named and no region choice, customer cloud or on premises option described.

Sourcekaaj.trust.siteread 2026-09-28

Prebuilt Agents / Templates / Packs Partial

Kaaj ships named agents and modules (document intelligence, business verification, bank statement analysis, credit memo automation, fraud detection) with segment solutions for equipment finance, brokers, merchant cash advance and community banks, but the agents are stages of one underwriting pipeline rather than 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, so a package forwarded to the lender's Kaaj address, submitted through an embedded form or posted to the API starts the workflow with no one at the lender asking, and webhooks report status changes.

Sourcekaaj.airead 2026-09-28

Model Flexibility & Routing Not documented

No model provider is named on Kaaj's pages or trust center subprocessor list and no customer model choice is described.

Sourcekaaj.trust.siteread 2026-09-28

APIs / SDKs / MCP Extensibility Partial

The homepage names a REST endpoint (POST /v1/applications) with webhook subscriptions and the FAQs say API and webhook workflows are supported, but no API reference, authentication scheme or access path to one is published.

Sourcekaaj.airead 2026-09-28

Testing, Debugging & Optimization Not documented

No way for a lender to test or score the agents' extractions, analyses or memos is documented; analyst edits and overrides are recorded, and the regression testing listed in the trust center guards Kaaj's own software.

Sourcekaaj.ai/productread 2026-09-28

Browser / Computer-use Not documented

Documents are parsed, public records checked and results synced through integrations; no browser, desktop or computer control is described.

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.

Agentic Index verified 2026-09-28

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

Head to head

Contact us

Found a vendor we missed? Have feedback on the index? We'd love to hear from you.