AI Research Automation with Perplexity

Automate research questions with Perplexity. A Jinba Flow workflow sends each question to Perplexity, checks the answer against your evidence rules and files what meets your threshold.

Book a Demo
  1. Research question raised An analyst needs an answer
  2. Query Perplexity Answers returned with their sources
  3. Assess the sources Check the answer against your evidence rules
  4. Evidence threshold Weak sourcing routes to a human
  5. File the answer Answer and sources kept together

What you can do with Perplexity + Jinba

  • Auditable AI Search in Workflows

    Incorporate Perplexity AI search and reasoning into Jinba Flow workflows with full audit logging on every execution, giving compliance teams a complete record of how AI outputs were generated and used.

  • Queries You Control

    Jinba Flow runs on-premise or in a private cloud, while each Perplexity step sends its query to Perplexity. Write queries that name the topic, not the customer, and the run history shows each one.

  • Team-Wide Workflow Governance

    Publish Perplexity-powered workflows as shared team assets in Jinba Flow with role-based permissions, SSO, and Active Directory integration so only authorized staff can build, modify, or execute approved automations.

  • Deterministic Workflow Execution

    Wrap Perplexity AI calls inside Jinba's 80% rule-based deterministic architecture, producing consistent and auditable outputs at scale while reducing the token costs associated with fully stochastic AI agent approaches.

How teams use it

KYC Document Review with AI Search

Automate KYC intake by routing identity and compliance documents through Perplexity AI search inside a governed Jinba Flow workflow, with every decision step logged for regulatory audit and team-level access controls enforced.

Compliance Research and Reporting

Use Perplexity AI to surface relevant regulatory information within a deterministic Jinba workflow, then generate structured compliance reports that are version-controlled, auditable, and publishable as reusable team APIs or batch processes.

Contract Review Automation

Embed Perplexity AI reasoning into contract review workflows built in Jinba Flow, enabling legal and operations teams to process documents consistently under RBAC-enforced governance with full audit trails.

Loan Underwriting Research Automation

Integrate Perplexity AI search into loan underwriting workflows to gather and synthesize relevant financial data, while Jinba's deterministic execution layer ensures every step remains auditable and compliant with banking regulations.

Learn how

How to Research with Perplexity in a Jinba Workflow

Retrieve, summarise and route a research answer in three steps, each written to the same audit log.

Works well with

Microsoft SharePointMicrosoft 365

Automate document intake, extraction and routing

AWS Bedrock Knowledge BaseSearch & Data

RAG and document sync inside your own VPC

Microsoft Entra IDMicrosoft 365

Look up M365 profiles and manager relationships

OpenAIAI Models

Call OpenAI models from any flow

Anthropic (Claude)AI Models

Invoke Claude models as a workflow step

Frequently Asked Questions

Everything you need to know about Jinba. Can't find the answer you're looking for? Reach out to our support team.

Does Jinba support Perplexity integration in an on-premise environment?

Jinba Flow can run on-premise or in a private cloud. Perplexity steps send their query to Perplexity's API, so write queries that name what you are researching without pasting customer records. An air-gapped deployment cannot reach Perplexity.

Is my data sent to a public AI model when using Perplexity with Jinba?

Yes, the query in a Perplexity step goes to Perplexity. Other steps can use private model hosting through AWS Bedrock, Azure AI or a self-hosted model, and the run history shows exactly what each step sent.

Does the Perplexity integration in Jinba produce audit logs for compliance?

Yes. Audit logging is a core built-in control in Jinba Flow. Every workflow execution that includes a Perplexity AI step is logged, giving compliance officers and auditors a complete, timestamped record of inputs, outputs, and decision steps. This satisfies the auditability requirements common in regulated industries such as banking, insurance, legal, and healthcare.

Can non-technical staff run Perplexity-powered workflows safely?

Yes. Workflows built in Jinba Flow using the Perplexity integration can be published to Jinba App, where non-technical business users such as compliance analysts, KYC processors, or loan officers execute them via a conversational chat interface with auto-generated input forms. Role-based permissions (RBAC) ensure that only authorized personnel can access and run each workflow, separating the building of workflows from their execution.

How does Jinba reduce AI token costs when using Perplexity at scale?

Jinba's architecture is 80% rule-based and deterministic, meaning AI model calls like those to Perplexity are invoked only at the specific workflow steps that require them rather than on every execution pass. This targeted use of AI within a deterministic workflow structure significantly reduces token consumption compared to fully stochastic AI agent approaches, addressing the CFO-level concern around enterprise AI API spend at production scale.

Ready to Govern Your Perplexity Workflows

Talk to our team about building secure, auditable Perplexity AI workflows that meet your regulatory requirements.

Book a Demo