Ananya Dhangar

Built an AI loan underwriting system, iteratively optimizing LLM prompts to eliminate hallucination and enforce output consistency across CIBIL reports, PD interviews, valuation, and legal summaries — cutting manual credit review by 50% and decision latency by 20%. Developed hierarchical dashboards and on-demand reporting systems providing 350+ agents across 80+ branches real-time visibility into field performance, incentive targets, and collection health. Optimized MongoDB pipelines with indexing and caching; background processing with Celery. Implemented loan document automation (welcome letters, e-stamping, e-signing, and SOA ledgers) reducing loan turnaround by 40%. Built a bulk WhatsApp notices engine dispatching 5k–6k notices monthly across escalation stages. Field visit optimization via geo-spatial clustering to auto-assign 500+ field officers to shortest routes, cutting travel costs by 25%. Document intelligence pipeline using Amazon Rekognition and OpenCV to mask PAN, Aadh

Developed a RESTful API platform integrating multiple HRMS systems using Django REST Framework, Docker, and AWS S3 for scalable enterprise data sharing. Implemented Redis caching reducing API response times by 30% and Celery async processing increasing system throughput by 20%. Designed a scheduler-driven change log tracking 100K+ daily data points, eliminating 100% of manual tracking effort.