Data Analyst, Collections & Lending Analytics (Delivery)

Mumbai

2 - 4 years experience

Responsibilities

About Gnani.ai and the team

Gnani.ai builds proprietary voice and conversational AI – ASR (Automatic Speech Recognition), TTS (Text-to-Speech), SLMs (Small Language Models) and agentic voice platforms such as Inya.ai – across 22+ Indian languages.

The Delivery team runs AI voice and messaging campaigns for leading Indian banks, NBFCs (Non-Banking Financial Companies) and retail lenders.

Collections portfolios: Personal Loan, Business Loan, Auto Loan and Gold Loan, across Pre-due, Post-due (Self-cure), Charges Collection and Settlement buckets.

Sales portfolios

lead qualification for SME (Small and Medium Enterprise) loans, Top-up loans and Refinance/Balance Transfer.

You will sit at the center of this engine: your segmentation and analysis decide who gets called, when, and with what message – and the insights you produce decide how much our clients collect and convert.

Role purpose

This role exists to turn campaign, call-outcome and payment data into strategies that measurably lift collection efficiency and lead conversion for our clients. File uploads and campaign operations sit with a separate owner; this role is purely analytical.

Insight

clients see clear, trusted dashboards and understand why numbers moved, not just that they moved.

Uplift: tested calling and messaging strategies that move connectivity, PTP (Promise to Pay) kept rate and collection efficiency month on month.

Proof

every claimed improvement is measured against a control group, so clients trust the impact attributed to Gnani.

Account growth

data-backed evidence that helps Delivery expand scope and revenue within each account.

Key responsibilities

Data quality and segmentation logic

Define segmentation rules (bucket, DPD – Days Past Due, ticket size, language, past behaviour) that the campaign operations owner executes.

Specify suppression logic – DND (Do Not Disturb), opt-outs, paid-customer exclusion – and audit that it is applied correctly.

Audit data quality weekly

dialed vs. allocated base, disposition consistency, payment-file match rates; flag gaps to the Project Manager.

Own the data dictionary for each account so metrics are calculated the same way every time.

Data logic and analysis

Build the full funnel per account: dialed connected RPC (Right Party Contact) engaged PTP paid.

Find drivers of drop-off by segment, time slot, language, retry count and script variant.

Use Mongo and Python to analyze large call and payment datasets and automate recurring analysis.

Build simple propensity views (who is likely to pay or respond) to prioritize calling effort.

Dashboards and reporting

Build and maintain client dashboards (Power BI or Excel) for daily, weekly and monthly reviews.

Present data in a clear, client-ready format: trends, bucket-wise and product-wise performance, and next actions.

Prepare data packs for QBRs (Quarterly Business Reviews) with the Project Manager.

Calling and messaging strategy

Design segment-wise strategies: time-of-day slots, retry cadence, language, channel mix (voice, WhatsApp, SMS – Short Message Service) and script variants.

Run A/B tests and measure uplift against a control group before scaling a change.

Recommend bucket-specific approaches (e.g., reminder-led for Pre-due, PTP-led for Self-cure, offer-led for Settlement).

Client communication

Lead analytical conversations with client collections, risk and product teams: performance reviews, strategy proposals, test readouts.

Explain numbers in business terms (amount recovered, roll-back avoided) rather than call counts.

Manage multiple accounts in parallel with clear prioritization agreed with the Project Manager.

Data protection and compliance

Handle borrower data (names, phone numbers, loan account numbers, outstanding amounts) as sensitive PII (Personally Identifiable Information).

Follow the DPDP Act (Digital Personal Data Protection Act, 2023), RBI (Reserve Bank of India) Fair Practices Code and recovery-agent guidelines, and TRAI (Telecom Regulatory Authority of India) calling-hour and DND rules.

Use only approved, access-controlled storage; never share client data over personal email or messaging apps.

Must-have skills and qualifications

Bachelor's degree in Statistics, Mathematics, Economics, Engineering, Commerce or a related field.

2–4 years in collections analytics, lending/risk analytics or campaign analytics at a bank, NBFC, fintech or collections agency or running campaigns for an Indian Voice AI Startup

Advanced Excel

XLOOKUP/VLOOKUP, INDEX-MATCH, pivot tables, IF/IFS, COUNTIFS/SUMIFS, Power Query.

Strong Database Skills

JOINs, GROUP BY, CASE WHEN, CTEs (Common Table Expressions), window functions.

Working Python (pandas) for analysis and automating recurring reports.

Applied statistics

control groups, significance testing, uplift measurement.

Dashboarding in Power BI with clear visual storytelling for business audiences.

Preferred

Working knowledge of collections metrics and bucket logic (see Domain knowledge).

Clear written and spoken English; Hindi or another Indian language is a plus for client calls.

Good-to-have skills

Experience with dialer, CRM (Customer Relationship Management) or LMS (Loan Management System) data exports.

Propensity or risk scoring (e.g., logistic regression, decision trees) to prioritize who to call and when.

Exposure to voice AI or contact-center analytics (call dispositions, AHT – Average Handle Time, sentiment).

Workflow automation

scheduled scripts, SFTP (Secure File Transfer Protocol) or API (Application Programming Interface) based file transfers.

Lead-funnel analytics for loan sales: lead-to-qualified, qualified-to-disbursal conversion.

Skills Required

Primary Skills

Advanced Excel (Pivot Tables, VBA)

VLOOKUP

Advanced SQL (CTEs, Window Functions)

JOINS and FUNCIONS

Pivot Tables

Pivot Table

XLOOKUP