Disclaimer: Independent product / portfolio work by Kaushal Khodifad. Not a live commercial product.return to portfolio
Disclaimer: Independent product / portfolio work by Kaushal Khodifad.
For Lyzr · PM - BFSI Vertical

Why I built ResolveIQ

The role asks three questions and says “show us you can do it.” Here are the answers - each a one-line claim, then the proof rendered as a diagram, a comparison, or a plan - with a working console behind them.

1 · What would you build on Lyzr for BFSI?

ResolveIQ - autonomous payment-dispute & chargeback resolution, as a new pillar in Agent Amadeo. Six specialized agents under a manager, sequenced by deterministic HybridFlow so the Reg E & network-deadline logic is code, not a hopeful prompt. It even refuses to act on a $1,310 friendly-fraud case and hands it to a human.

Dispute Manager
Orchestrator
AgentMeshSLA TimerRBAC
fans out to
Audit & Reporting
Seals every step
Audit LedgerExplainability Layer
Intake & Triage
Salesforce CRMEmail/Chat connectorPII Redaction
Evidence Gatherer
MeridianCore APIVROL / Mastercom recordsGroundedness
Fraud Detection
Fraud FeedOFAC ScreenFairness & Bias Manager
Risk Scoring
HybridFlow (structured ML)Reflection Mechanism
Policy & Decision
Rules EnginePolicy EngineGroundedness
Settlement & Filing
Meridian GLVROL / Mastercom filingOutput Guardrail
Managerial pattern · AgentMeshDeadline & policy logic = deterministic HybridFlowDeployable to bank VPC / on-premHover an agent to see what it touches

2 · Why is it the right use case?

Because disputes are where agents win the argument outright. A ~$15B, double-digit-growth problem with hard deadlines and evidence across five systems - irreducibly multi-step, deadline-bound, and human-in-the-loop. RPA breaks; a chatbot can't act. And it isn't an orphan bet: Lyzr published the dispute playbook + six-agent architecture, and Accenture invested to push agentic AI into banking.

Can it…RPAChatbotAgent
Reason on ambiguous evidence
Act across 5 systems
Hold a Reg E / network deadline
Leave a replayable audit trail

3 · How would you position it & take it to market?

Not as another agent builder - as the production-and-compliance layer for BFSI agents. Everyone can demo a dispute agent; ResolveIQ ships because explainability, tiered autonomy, and data sovereignty are native - built so the CRO and Model-Risk sign off instead of block. That's the exact gap where ~95% of GenAI pilots stall (MIT 2025) and 40%+ of agentic projects get cancelled by 2027 (Gartner). Five moves:

1
Ride the Accenture channel
9-12mo → faster

Co-sell into Accenture's BFSI accounts where the MSA & trust already exist. Neutralizes “will you exist in 3 years?”

2
Paid, value-gated POC
≥85% on 500 cases

Runs in the bank's VPC on masked data. Exit gate = measured result + named sponsor + budget line.

3
Sell the P&L, meter the runs
$37 → ~$9 / dispute

Platform + VPC usage ($0.03/run), then outcome pricing per dispute auto-resolved. Framed to the COO, never as “agent runs.”

4
Ship a ready Trust Pack
SOC2 · ISO · RBI/SR11-7

VPC/on-prem, “we don't train on your data,” audit-replay spec + Responsible-AI one-pager. In BFSI, blockers stop more deals than buyers start.

5
Land then expand
NRR > logo count

By use case (disputes → KYC → AML → loan servicing) and by BU (retail → SME → wealth, or insurance claims). Same plumbing.

The asset is live - not a deck.

Run a real dispute through the six agents, watch it auto-resolve, and watch it stop for a human.

Open the console
sri@lyzr.aividur@lyzr.ai· by Kaushal Khodifad

↑ Independent product concept by Kaushal Khodifad. ResolveIQ is an independent product concept by Kaushal Khodifad; it is not a real company or a commercial product. It explores the autonomous payment-dispute resolution on Lyzr space. Not a live commercial product. Data is illustrative.

Return to portfolioOpen the live demo