Our Work

Engagements across Logistics, Insurance (Mutual Funds), Media, Mobility, Consulting and Technology — building the data foundation first by integrating varied data sources, then the intelligence layer that business leverages to make smart decisions.

Business Intelligence Platform for a Logistics Client (Pilot Implementation)

Client Profile: A technology-enabled mobility/logistics operator managing a large EV fleet deployed across multiple ride-hailing and mobility platforms such as Uber, Ola, Rapido and Namma Yatri

Business Challenge: The client had operational data across multiple platforms, vehicle/fleet records, driver payouts and finance data, but management lacked a unified view of vehicle-level and fleet-level Operational and Financial insights.

Solution Provided: Built Financial intelligence platform creating a single financial and operational intelligence layer across fleet, trip, driver and finance data. Raw vehicle data feeded to data platform, transformed & aggregation on business KPI’s.

Medallion data architecture adoption enabled three stages of data processing from Raw – staging – Curated (Bronze/Silve/Gold stages). Vehicle Telematics data integration capturing real time mobility data along with vehicle performance. Vehicle Trip data fed from aggragators (Uber/Ola/Rapido)

Technology: PySpark, AWS S3, Apache Iceberg, Power BI Reporting/Dashboards

Business Outcome:

Clear financial visibility and revenue Analytics

Fleet Utilization analytics providing accurate operational performance

Driver Economics (Measure, Track & Optimize)

Cost Analytics

Profitability

Forecasting to maintain realistic view of growth trends

AP Automation (Document AI) - Mutual Fund Service Provider

Client Profile: An investment company offers secure mutual fund services in India. Simplify investments, track portfolios, and support mutual fund distributors with digital tools.

Business Challenge: Client currently handles Vendor invoices thru manual processing validating the invoice and PO details based on delivered items by the vendor. The client receives over 3000 invoices a month and it takes huge effort towards validation, approval/rejection workflow, payments tracking and Invoice issue resolution etc.,

Main issues requiring support involved: Periodic invoice tracking and clearance of payment, Information to vendors on rejected invoices and follow up of revised invoices. Invoice validation is cumbersome as several mismatches to Purchase orders (PO’s) causing payment rejection.

Solution Offered: Automated invoice scanning from PDF files collected from central File folder and applied Document AI based data capture algorithm and run validation using business rules against purchase orders extracted from finance system. Post validation the approval workflow process gets initiated enabling supervisor to approve matched invoices and unmatched ones will be sent for rejection to supervisor review.

Vendors will intimated thru emails communication on invoice mismatch with PO and relevant correction will be requested. Consolidated invoice data report gets generated with list of approved/mismatch (reject) invoices for historic and audit trail purposes.

Technology: Python, pdfplumber, Tesseract OCR, PaddleOCR, NLP, MySQL DB and Jason

Business Outcome:

Automation of Invoice – Purchase Order process eliminating the manual validation

Significant Cost/effort savings per month (2 HC)

Timely completion with no delay/bottlenecks and faster payment processing

Solution caters to handle additional volume of invoices in future