Home Solutions ↳ AI Cost Engineering ↳ Training & Enablement Why Us Team Case Studies CopilotFactory ↳ ClaimFlow AI ↳ Milo AI Blog Contact
Product Spotlight · ClaimFlow AI

AI-Driven Claims Intake Engine

Bridging the gap between customer interaction and backend data readiness — ending the era of manual data intake in insurance claims processing.

Try Now
45%
Faster claim registration
70%
Less manual data entry
4
Fields captured conversationally
2
Azure App Services, both healthy
The Problem

Traditional claims intake
is slow and manual.

Customers call a contact center, wait for an agent, and answer the same repetitive questions every time — while the agent manually keys everything in.

Manual data entry introduces data quality issues, and every extra minute on the call compounds directly into higher operational cost.

See the Solution
01

Long Wait Times

Customers call in and wait for an available contact center agent before intake even begins.

02

Repetitive Questions

The same set of intake questions are asked manually, claim after claim.

03

Data Quality Issues

Manual entry introduces inconsistent formatting and field-level errors.

04

Rising Operational Cost

Longer average handling time per claim adds up directly to cost.

Enterprise Deployment

Deployed for AIG — at enterprise scale

ClaimFlow AI went from scoping to production inside AIG's claims environment in six weeks — integrated directly into their FNOL workflow, not bolted on top of it.

Workflow Integration

ClaimFlow connected to AIG's backend claims platform via secure API connectors — reading policy records in real time and pushing completed FNOL records straight into the queue.

  • Policy validation against AIG's live policy database
  • Real-time FNOL submission to AIG's claims backend
  • Full per-conversation audit trail retained for compliance

Phased Rollout

Deployment followed a gated rollout — each phase unlocked only after quality metrics cleared, keeping AIG's live claims traffic fully protected throughout.

  • Phase 1: Web-channel conversational intake (Weeks 1–3)
  • Phase 2: Contact-centre overlay in agent-assist mode (Weeks 4–5)
  • Phase 3: Voice-enabled intake via Azure Speech (Week 6)

ROI at AIG

Measured against AIG's pre-deployment baseline across the same FNOL volume, results aligned with CodeSizzler's enterprise benchmarks.

  • 45% faster claim registration end-to-end
  • 70% reduction in manual data entry per claim
  • 30–40% operating cost savings on intake headcount
  • 2–3× claims throughput on the same team footprint
The Solution

One conversation.
A structured claim record.

ClaimFlow AI runs a state-machine-driven conversation that collects, validates, and extracts every required field — without a single form.

Policy Number

Agent: "Hello, please provide your 8-digit policy number."

policy_number12345678

Incident Timestamp

Agent: "What is the date and time of the incident?"

incident_timestamp2025-06-18 10:30 AM

Location

Agent: "Where did the incident occur?"

locationDallas, Texas

Damage Description

Agent: "Can you briefly describe the damage?"

damage_descriptionRear-end collision at traffic signal

Compile & Submit

All four fields compiled into one structured claim record and submitted.

policy_number12345678 ✓
incident_timestamp2025-06-18 10:30 AM ✓
locationDallas, Texas ✓
damage_descriptionRear-end collision ✓

State Machine

START → POLICY_NUMBER → INCIDENT_TIME → LOCATION → DAMAGE_DESCRIPTION → SUBMIT. The agent always knows exactly which question to ask next.

Current Status

Live on Azure App Service

Both services are deployed and healthy, running independently on Azure.

● Healthy

ClaimFlow-AI — Backend

FastAPI service running on Azure App Service, handling conversation orchestration and claim extraction.

● Healthy

ClaimFlow-UI — Frontend

Streamlit interface running on Azure App Service, deployed via a dedicated GitHub Actions workflow.

What's Next

On the roadmap

01

Automatic Voice Output

Full text-to-speech playback after every agent response, so the conversation feels truly hands-free — speak, listen, respond, repeat.

02

Custom Domains

claimflow.codesizzler.in for the customer-facing UI, and api.claimflow.codesizzler.in for the backend service.

03

Full Production Demo Flow

Voice input → Azure Speech-to-Text → ClaimFlow AI Agent → field extraction → live audit panel → Azure Text-to-Speech → voice response → claim submission.

Product Spotlight

VoiceRoute Kitchen — Powered by Milo

A conversational ordering agent that greets, guides, and checks out your customer — by voice or text — without a single form or human operator.

Try Now
45
Built capabilities, shipped and working
3+
Languages understood (EN / HI / TA)
Live Razorpay checkout with demo mode
0
Forms filled by the customer — ever
The Problem

Restaurant ordering is
still clunky and impersonal.

Customers scroll through static menus, tap confusing UI flows, and never get to say "make it Jain" or "extra gravy, please" in plain language.

Operators lose upsell moments, dietary notes get missed, and every edge case — a nut allergy, a last-minute swap — becomes a support call. The gap between what a customer wants and what the kitchen gets is wide and expensive.

See the Solution
01

Static Menus Miss Nuance

No form field for "no onion, no garlic, extra gravy" — customizations fall through or get mis-keyed.

02

Dietary Errors Are Costly

Nut allergies, Jain requirements, and vegan needs often go untagged until the food reaches the table.

03

Missed Upsell Moments

A static UI never suggests the naan that pairs perfectly with butter chicken — Milo does, once, at the right moment.

04

Language Barriers

Indian restaurants serve multilingual customers. A single-language UI leaves a significant segment underserved.

The Solution

One conversation.
A confirmed kitchen order.

Milo runs a natural, memory-aware conversation that captures every detail — dining mode, items, spice, customizations, dietary flags — and closes with a real payment.

Greet & Route

Milo greets the customer and asks: Dine-in, Takeaway, or Delivery?

dining_modedine-in

Take Order Item-by-Item

Each item: name, spice level, customizations. All tracked across the full conversation.

itemButter Chicken
spiceMedium
customextra gravy

Dietary Flag & Upsell

Jain / allergy tags are set automatically. One complementary suggestion offered — never repeated if declined.

tags[Jain]
upsellGarlic Naan?

Read Back & Confirm

Full order read back grouped by category — starters, mains, drinks — with itemized ₹ totals including GST.

subtotal₹680
gst_9pct₹61
total₹741

Checkout & Pay

Dedicated checkout page: Cash / Counter / Razorpay online. Demo mode auto-activates when live keys are absent.

order_idSRK-17198423
paymentrazorpay
statuspaid ✓

Conversation State

GREET → DINING_MODE → ITEM_LOOP → DIETARY_CHECK → UPSELL → CONFIRM → CHECKOUT → THANK_YOU. Milo always knows exactly where in the order it is.

What Milo Can Do

45 capabilities, fully built.

Everything from voice input to live payment — shipped and working.

Conversation & Ordering
Greets and asks dine-in, takeaway, or delivery
Takes orders item-by-item by text or voice
Asks spice level per item — mild to extra spicy
Accepts free-text customizations per item
Tracks all items in memory across the full conversation
Suggests one complementary item — once, never repeats
Supports mid-conversation add, remove, or swap
Reads back full order by category before confirming
Emits structured ORDER JSON into the live order panel
Dietary & Safety
Tags Jain requests ("no onion no garlic") with [Jain] in the order
Guides vegan questions to dairy-free veg dishes
Green / red dot indicators for Veg vs Non-Veg throughout
Language
Replies in English by default
Understands and replies in Hindi via GPT-4o
Understands Tamil input via GPT-4o + Azure STT
Voice Interface
Press-and-hold mic records WAV, sent to Azure STT
TTS auto-plays after every Milo response
Speech rate boosted ~15% above default
Markdown symbols and emojis stripped before speaking
Line breaks converted to SSML pauses for natural pacing
Falls back to type-instead prompt on STT failure
Order Summary
Read-back grouped by category — starters, mains, drinks
One item per line in chat — no run-on text
Pricing
All prices in ₹ Indian Rupees
GST at 9% on subtotal, rounded to nearest rupee
Subtotal, GST, and grand total shown separately at all times
UI & Frontend
Centered mobile-phone-frame layout
Live Order Panel updates in real time as items are added
Gold/maroon Indian-heritage theme — filigree and paisley motifs
"View Menu" opens a full categorized menu modal
Custom SVG restaurant logo for Milo
Animated typing indicator while Milo thinks
Mic button with idle, recording, and processing states
Checkout & Payment
Dedicated full-screen checkout page after confirmation
Cash on Delivery, Pay at Counter, or Pay Online Now
Razorpay integration for live online payments
Demo/mock payment mode auto-activates without real keys
Thank You page with Order ID, badge, and Place New Order reset
Backend & Order Handling
Generates unique Order ID (SRK- + timestamp)
Logs full order details server-side on submission
Returns confirmation and estimated prep time to frontend

Tech Stack

GPT-4o Azure OpenAI Azure Speech Services en-IN-NeerjaNeural TTS FastAPI Uvicorn Python Razorpay SDK menu.json API SSML WAV Audio Pipeline GitHub Actions
Current Status

Live & Actively Developed

Core ordering, voice, and payment are all live. A few edge-case features are pending verification or are paused on external dependencies.

● Working

Core Conversation & Ordering

All 9 ordering capabilities, dietary tagging, upsell logic, order JSON output, and Razorpay checkout are live and tested end-to-end.

⚠ Needs Verification

Nut Allergy Flag & Vegan Handling

Nut allergy flag and vegan-option routing didn't appear in the last tested order JSON — needs explicit test coverage and a fix if missing.

⚠ Known Limitation

Regional Language TTS

Hindi and Tamil text responses are spoken through an English-accented voice engine. No true regional STT locale or matching native TTS voice is configured yet.

⏸ Paused

OTP Login & SMS / Email Summary

Phone/email OTP login and order-summary delivery via SMS or email are paused — blocked on a Twilio outage and not yet built.

What's Next

On the roadmap

01

True Regional Language TTS

Configure a dedicated Hindi and Tamil STT locale and matching native TTS voice (e.g. hi-IN-SwaraNeural, ta-IN-PallaviNeural) so the voice experience matches the customer's language from end to end.

02

OTP Login + Order Summary Delivery

Phone/email OTP login at the start of each session, followed by an SMS or email order summary after checkout — currently paused on Twilio outage resolution.

03

Nut Allergy & Vegan Flag Verification

Explicit test suite to confirm allergy flags and vegan routing appear correctly in the output JSON for every edge case, with a fix shipped if missing.

04

Kitchen Display Integration

Push confirmed order JSON directly to a kitchen display system or POS via a webhook, removing the final human relay between Milo's output and the kitchen.