Overview

Product tour: AI call QA and compliance in two minutes
Two-minute product tour with illustrative data: a Hinglish loan call analysed, an evidence-backed QA scorecard, compliance checks, dashboards and the architecture that keeps call data in your account.
Product tour: AI call QA and compliance in two minutes
AI call analysis for every recorded call
Evidence-backed call QA scorecard
Automated call compliance checks
Contact centre overview dashboard
Agent performance and coaching dashboard
Buying signals and missed leads
Churn risk and customer journey
Searchable call list
Runs in your AWS account
Plans at a glance
Ozoo Post Call Analytics is AI post-call analytics for contact centres in banking, financial services and insurance (BFSI). It analyses every recorded call, in Hindi, Hinglish and English, and turns it into a transcript, a summary, a disposition, a quality score with evidence and compliance flags, inside your own AWS account.
Who it is for
Banks, NBFCs, housing-finance companies, insurers, fintech lenders and collections agencies in India running sales, service and collections calls, and the QA, compliance, training and operations teams that review them.
What every call gets
- Transcript. Agent and customer separated, code-switched Hindi and English handled, your product terms boosted. Click any line to play the audio from that moment.
- Summary and disposition. A plain-language summary, topic, product, resolution and a disposition from your own category and sub-category taxonomy.
- QA scorecard with evidence. The AI answers each question on your scorecard and cites the transcript; code computes the score; fatal errors are re-checked by a stricter second pass. QA analysts can override a mark, and every override is written to an audit trail.
- Compliance checks. Fatal and critical rules you define, such as mis-selling, misleading information or undue pressure, and the rates, fees and offers quoted on the call matched to your product catalogue.
- Customer and revenue signals. Sentiment, a customer-experience score, churn risk, buying signals and missed leads, and episodes that group repeat calls about one issue.
Dashboards and workflow
Overview, coaching, growth and customer-success dashboards; agent leaderboards and coaching profiles; a searchable call list with filters; reviewer comments; export to Excel; agent and team management with bulk import; role-based access. Prompts, scorecards and dispositions are configuration you edit in the application, not code.
Architecture and data residency
- Deployed by AWS CloudFormation into your AWS account in Asia Pacific (Mumbai). HabileLabs has no access to your recordings, transcripts or results.
- Your data stays in your account, in Amazon S3 and Amazon Aurora PostgreSQL, encrypted at rest and in transit. Summaries, classification and scoring run on Amazon Bedrock in your account.
- Speech-to-text by Deepgram (Nova-3) with your own API key; Sarvam and AssemblyAI are alternatives. Call audio leaves your account only for transcription; the daily licence check-in sends usage counts, never call content.
- Serverless: AWS Lambda, AWS Step Functions, Amazon SQS and Amazon API Gateway.
Pricing
Plans differ only in analysed call minutes a month: Starter 25,000, Growth 100,000, Business 500,000, Enterprise 2,000,000. Add 25,000-minute packs; minutes beyond your allowance are billed per minute. Month-to-month or 12, 24 or 36-month terms, or a private offer for INR billing, a site licence or a 30-day trial. AWS infrastructure and your speech-to-text provider are billed to you separately.
Frequently asked questions
What is post-call analytics? Analysing recorded calls after they end: transcription, summaries, classification, quality scoring and compliance checks on every call rather than a sample.
Does it support Hindi and Hinglish calls? Yes. Hindi, English and code-switched Hinglish calls are transcribed and analysed, with results in English.
Do we need Amazon Connect? No. Any telephony system or dialler that can write recordings to Amazon S3 works, and users can upload recordings in the application.
Where is our call data stored? In your AWS account in Mumbai. Only the audio goes to your speech-to-text provider, for transcription.
Is it real-time? No. Calls are analysed after they end, typically within minutes.
How is it deployed, and how long does it take? Subscribe here, create your account on the HabileLabs Marketplace portal and launch the CloudFormation template with the licence key it gives you. The stack builds its own network, database, storage and application in under an hour; most teams set up scorecards, dispositions and agents within a week.
What do we need first? An AWS account with access to Anthropic Claude models in Amazon Bedrock in Mumbai, and a Deepgram, Sarvam or AssemblyAI API key.
What support is included? Email support Monday to Friday, 10:00 to 19:00 IST, plus implementation, scorecard design and integration services.
Highlights
- Review 100% of calls, not a 2-5% sample: speaker-separated Hindi, Hinglish and English transcripts, AI summaries, dispositions, sentiment and a QA scorecard where every mark cites the moment in the call behind it.
- Compliance on every call: your fatal and critical rules checked automatically, fatal errors re-checked by a stricter second pass, and the rates, fees and offers an agent quotes matched to your product catalogue.
- Runs in your AWS account in Mumbai: recordings, transcripts and results stay in your Amazon S3 and Aurora PostgreSQL, and the analysis runs on Amazon Bedrock in your account. Deployed with AWS CloudFormation.
Details
Features and programs
Financing for AWS Marketplace purchases
Quick Launch
Pricing
Free trial
Dimension | Description | Cost/month |
|---|---|---|
Starter plan | 25,000 call minutes a month. All features, production and non-prod installs. | $400.00 |
Growth plan | 100,000 call minutes a month. All features, production and non-prod installs. | $1,200.00 |
Business plan | 500,000 call minutes a month. All features, production and non-prod installs. | $4,000.00 |
Enterprise plan | 2,000,000 call minutes a month. All features, production and non-prod installs. | $12,000.00 |
25,000 additional call minutes | Adds 25,000 analysed call minutes a month to any plan for the contract term. | $350.00 |
The following dimensions are not included in the contract terms, which will be charged based on your usage.
Dimension | Description | Cost/unit |
|---|---|---|
Call minutes beyond your allowance | Each call minute analysed beyond the plan and pack allowance in a calendar month. | $0.015 |
Dimensions summary
Top-of-mind questions for buyers
Vendor refund policy
To request a refund, contact our support team within 7 days of the billing date with relevant details. Approved refunds will be processed through AWS Marketplace as per their policies.
How can we make this page better?
Legal
Vendor terms and conditions
Content disclaimer
Delivery details
Software as a Service (SaaS)
SaaS delivers cloud-based software applications directly to customers over the internet. You can access these applications through a subscription model. You will pay recurring monthly usage fees through your AWS bill, while AWS handles deployment and infrastructure management, ensuring scalability, reliability, and seamless integration with other AWS services.
Additional details
Usage instructions
Deploy Ozoo Post Call Analytics in your AWS account
Ozoo Post Call Analytics runs entirely in your own AWS account, in Asia Pacific (Mumbai). Installation takes about 40 minutes and needs permission to create the resources in the IAM policy shown next to the template. The deployment and data handling guide covers everything it creates.
1. Before you start
- Amazon Bedrock access to Anthropic Claude Sonnet 4.5 and Haiku 4.5. Anthropic models are sold through AWS Marketplace, so the account needs a valid payment method; submit the models' use-case details in the Bedrock console if asked.
- Room for one more VPC, internet gateway and Elastic IP address in ap-south-1, or two private subnets of your own VPC that reach the internet through a NAT gateway (set ExistingVpcId and ExistingPrivateSubnetIds). The stack's pre-flight checks these first and stops within a minute, saying what to fix.
- A speech-to-text API key from Deepgram (default), Sarvam or AssemblyAI. Call audio is sent to that provider for transcription under your account.
2. Create your HabileLabs Marketplace account
After subscribing you are redirected to the HabileLabs Marketplace portal . Register there; the portal shows your licence key (it starts with PCA-) and launch links for your production and non-production installs.
3. Launch the stack in Asia Pacific (Mumbai)
Use Quick Launch with the template below, or the portal's launch link, in ap-south-1. Fill in:
- Licence key: your PCA- key
- InstallSlot: production (or nonprod for a UAT stack; use a different stack name, for example ozoo-pca-nonprod)
- AdminEmail and AdminUsername: the first administrator; the sign-in invitation goes to this address
- TranscriptionEngine and TranscriptionApiKey: your speech-to-text provider and its key
- RetentionDays: how long recordings and analysis files are kept (default 1460)
- Optional ExistingVpcId and ExistingPrivateSubnetIds: install into your own VPC instead of a new one
Acknowledge that the template creates IAM resources and create the stack. A new account's Lambda concurrency limit is requested automatically when it is too low (see the Preflight output).
4. Sign in and send calls
Open WebAppURL from the stack outputs and sign in with the temporary password emailed to the administrator. Point your dialler or telephony platform at the input bucket and prefix in the stack outputs, or upload recordings from the call list, then set up categories, dispositions, scorecards and agents in Settings.
Upgrades and support
The install checks in with the portal daily; new releases appear in the portal with release notes and are applied by updating the stack with the new template. Support: info@habilelabs.io , Monday to Friday 10:00 to 19:00 IST.
Resources
Vendor resources
Support
Vendor support
HabileLabs supports Ozoo Post Call Analytics customers from subscription through deployment and day-to-day use.
Support includes:
Deployment assistance for the CloudFormation stack in your AWS account Scorecard, disposition, category and prompt configuration guidance User, role, agent and team set-up Recording ingestion and telephony integration guidance Technical issue resolution and troubleshooting Upgrades to new releases AWS Marketplace subscription and licensing questions
Support Contact: Email: info@habilelabs.io Website: https://www.ozoo.ai/products/post-call-analytics/ Support Availability: Monday to Friday, 10:00 AM to 7:00 PM IST, excluding public holidays.
Response Time: Our team typically responds within 24 business hours, depending on the severity and nature of the request.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.