Sanskar Goel

Delivered more than 75% of the backend for a very critical loan origination system within 2 weeks, with sole ownership of it.

Designing 50+ RESTful APIs, integrating seamlessly with frontend interfaces and internal services, and implementing core business and validation logic-enabling on-time launch despite last-minute scope changes and intense delivery pressure.

Engaged in design, development, integration, and resolving critical production issues and bugs, improving overall code quality and maintainability within 10+ Java-based backend services within a microservice architecture. Enhanced service configurability by eliminating hardcoded dependencies and implementing dynamic configurations, enabling a plug-and-play architecture that minimizes code changes for onboarding new partners and improves scalability and maintainability.

Designed and developed a scalable data access service in Java, enabling seamless retrieval of cross-schema data for multiple services, incorporating advanced search and filtering capabilities with MongoDB aggregation pipelines, optimized indexing for performance, and enhanced configurability through custom Spring Boot annotations.

Developed an automated Funnel Report generation system in Java for daily summary emails to NBFCs and banks using SMTP, aggregating loan application data across multiple schemas, extracting key insights, and delivering structured reports via scheduled cron jobs and on-demand API triggers.

Led database and legacy Java codebase cleanup initiative, improving code readability, optimizing queries, eliminating repetitive logic, ensuring maintainability across various stages of the software development lifecycle. Developed and deployed 5+ Python automation scripts to streamline daily operations, reducing manual effort and accelerating task execution across the team

Developed and contributed in various components of the Operational Sub-System, the backend fault tolerance solutions for global telecom clients like Vodafone (Europe), Optus (Australia), T-Fiber (Telangana) using Java and Spring Framework, following agile methodologies.

Automated duplicate alarm detection across 3,000+ alarms daily with Python, reducing manual intervention and delays.

Contributed to the development of Unified Correlation Analyzer and value packs for clients.

Started from scratch by performing a literature survey on existing problems and proposed solutions. Worked on training an AI-enabled model for disease identification in paddy leaf and in the process tried several models to obtain the highest level of accuracy.

Increased the accuracy further by performing an image enhancement process to remove the unwanted background from the sample data to achieve an accuracy of 98.34 percent.

Took the initiative to build an Android application, because almost 70 percent of mobile users are Android users. Used Android Studio and completed the entire documentation to prepare a journal.