India’s agentic AI for
regulated enterprises.
Vihaya is building an agentic decisioning platform for BFSI, healthcare, and telecom — designed around an immutable audit chain, citation-backed reasoning, and confidence-floor escalation. DPDP-aligned by design. The platform substrate is shipped and in-repo; the first paid pilots are the immediate roadmap.
Production AI in the world's largest enterprises.
Each engagement is one vertical wedge, 12 weeks, deployed into the customer's environment. The system stays theirs.
Claim adjudication, automated
End-to-end agentic decisioning for health claim adjudication. Retrieval-grounded against the insurer's policy wordings + IRDAI guidelines, citation-backed, IRDAI-aligned audit chain. Reference solution in the repo; the first pilots are open.
Unified customer experience
Intent-driven CX automation built for Indian telecom scale. Replace fragmented IVR + bot stacks with one agentic platform that personalises every interaction in English, Hindi, and the major regional languages. Roadmap solution.
Loan underwriting & AML
Agent for loan-file decisioning and AML alert review. Reads the borrower’s file, the policy bulletin, and the prior decision history. Output is one cited, escalation-protected decision with the audit trail RBI examiners expect. Roadmap solution.
What you commit to your regulators, not what's under the hood.
What follows is the behaviour of the platform — what your reviewers, compliance team, and auditors experience. The underlying engineering substrate is Vihaya's proprietary IP and shown to customers under NDA during the pilot kickoff.
Cited decisions
Every decision references the exact policy clauses, clinical guidelines, or transaction patterns the agent grounded on. Reviewable from cold storage years later — exactly what RBI and IRDAI examiners ask for.
Confidence-floor escalation
Below the configured threshold, cases route to your existing nurse reviewers, credit officers, or compliance team with our recommendation attached. No silent 100% automation.
DPDP-aligned by design
Append-only audit, configurable PII redaction in logs, data-fiduciary controls baked in. Notice + consent surfaces, purpose limitation, and breach-notification hooks for CERT-In reporting.
Quality gates that block deploys
Hand-curated golden datasets and CI-blocking regression tests. Drift below threshold → the deploy never reaches production. Your medical director or compliance head signs off on the bar.
Model-agnostic, India-aware
Use the foundation-model provider your security team has already approved. Multilingual — English, Hindi, and the major regional languages — built in from day one.
Deployed in your environment
Runs in your AWS Mumbai, GCP Mumbai, or Azure South India region. No SaaS egress. We embed; you run.
12-week paid pilot. We embed. The system stays yours.
Vihaya is not a SaaS. We deliver a deployed service into your environment, against your data, with your team in the loop from week one.
Discovery & access
Signed BAA / DPA. Sandbox credentials. Policy corpus delivered. SME interviews.
Ingest & integrate
Service deployed into your environment. Policies indexed. Intake and outbound wired into your systems. Eval dataset built with your SME.
Shadow run
Service runs against shadow or parallel traffic. Weekly review of decisions vs. your baseline. Threshold locked to your medical director's tolerance.
Handoff
Runbook. Reviewer training. Success-metrics report. Go / no-go decision on production phase.
What people ask before the first call.
Straight answers. No marketing varnish. If your question isn't here, send it.
What is Vihaya?
Vihaya is an India-built agentic AI decisioning platform for regulated enterprises — BFSI, healthcare, insurance, and telecom. The intended engagement shape is a 12-week paid pilot per customer per workflow, designed to take a large share of decisions touchless with the remainder audited, cited, and routed to human reviewers. Pre-revenue as of May 2026; the architecture is shipped in-repo and the first design-partner pilots are the immediate roadmap.
Is Vihaya DPDP Act 2023 compliant?
Vihaya's architecture is designed for the DPDP Act 2023. Append-only audit logging records every data-processing action with actor, purpose, and outcome — supporting the data-fiduciary obligations the Act requires. PII redaction in logs is configurable per tenant. Notice and consent surfaces are integration points the engagement wires into the customer's existing data-principal touchpoints. The audit chain supports the breach-notification timeline mandated under the Act and CERT-In's 6-hour reporting window.
Does Vihaya support RBI's IT outsourcing framework for banks?
Yes. Vihaya is deployed inside the bank's VPC (AWS Mumbai, GCP Mumbai, or Azure South India). There is no data egress to Vihaya's environment, which addresses the core localisation and exit-clause requirements of RBI's Master Direction on Outsourcing of IT Services. The append-only audit trail supports the RBI examiner expectation that every automated decision is reconstructable. We provide the standard outsourcing-MSA artefacts — exit-management plan, BCP/DR runbook, right-to-audit clause — alongside the pilot SOW.
How is Vihaya different from Distyl, Sierra, or other US AI agent platforms?
Distyl, Sierra, Cresta, and similar US-headquartered AI platforms are world-class for US enterprise buyers and have shipped real customer outcomes. Vihaya is pre-revenue and is being built in India for Indian regulated enterprises: DPDP-aligned by design, RBI / IRDAI / CERT-In aware, designed to deploy in Mumbai or Bengaluru cloud regions (AWS ap-south-1, GCP asia-south1, Azure South India), priced in INR per engagement, and able to operate in English, Hindi, and the major regional languages. The architectural shape is comparable; the buyer experience and the local-context positioning are the differentiators.
What industries does Vihaya serve in India?
Lead verticals: health insurance (claim adjudication, prior authorization), BFSI (loan underwriting, AML alert review, KYC enhancement), and telecom (customer experience routing in multiple Indian languages). The architecture extends to general insurance claims, NBFC underwriting, supply-chain root-cause analysis for manufacturers, and benefits-eligibility processing for government schemes like Ayushman Bharat / PMJAY.
What does a Vihaya pilot cost?
Pilots will be fixed-price 12-week engagements quoted in INR per engagement, after the discovery call. Vihaya is pre-revenue and does not publish a price range until the first design-partner engagements set the reference. Design-partner customers get materially better economics in exchange for shaping the product roadmap. Expect a draft SOW within five business days of the discovery call.
How does the confidence-floor escalation work?
Every decision the agent produces carries a model-reported confidence score between 0 and 1. Cases that score below the configured threshold (default 0.75) are routed to the customer's existing human reviewers — nurse reviewer for a health insurer, credit officer for a bank, compliance analyst for AML — with the agent's recommendation, rationale, and citations attached. This forces human-in-the-loop coverage regardless of the model's verdict.
Which LLMs does Vihaya use, and where do they run?
Vihaya is model-agnostic. The default is OpenAI gpt-4o. Anthropic Claude, Azure OpenAI (India regions), and Google Vertex (Gemini, Mumbai region) are first-class alternatives. Customers choose based on their existing vendor approvals — most Indian regulated enterprises today prefer Azure OpenAI India or Vertex Mumbai because of data-residency comfort. Swapping the model does not change the audit chain, eval gate, API contract, or escalation queue.
Where is Vihaya based?
Vihaya is built in India and serves Indian regulated enterprises as the primary market. The company is pre-revenue as of May 2026; the first paid pilots will set the public reference for engagements, pricing, and customer outcomes.
The fastest way to find out if this works on your book is to run a pilot.
30-minute discovery call. We’ll identify your highest-pain wedge and send a draft SOW within five business days.