Akash Yadav

Led the strategic design and development of an AI-driven RFQ management system, modernizing procurement workflows for manufacturers. Integrated advanced NLP and ML models to analyze vendor responses, generate negotiation strategies, and assist with content creation, directly contributing to components relevant for AI-based search and intelligence. Engineered an AI-based price negotiation module featuring real-time bidding and automated counter-offers, showcasing expertise in complex, data-driven system design. Collaborated on system integration with third-party data APIs to support real-time insights and AI-powered vendor ranking, demonstrating proficiency in data sources for advanced search and analytics.

Spearheaded the architecture, development, and scaling of a robust search system using Elasticsearch, achieving a throughput of 200 requests per second. Enhanced relevance ranking by designing and integrating a scoring system with calculated and custom boosts (MPS, UPS, SKS, CBS, CSS), experimenting with Normalized Weighted Multiplicative and Exponential scoring models. Optimized performance to reduce p90 latency from 1.4s to 100ms through database query and caching optimizations via Elasticsearch tuning, conducting extensive stress and load testing. Established comprehensive performance and error monitoring using centralized logging (Sentry, Elasticsearch Explain API) for proactive issue detection and resolution. Contributed to search infrastructure optimization for low latency and high throughput, streamlining CI/CD pipelines, and implementing liveness/readiness APIs for system health checks. Integrated third-party tools like GrowthBook for feature flagging and A/B testing to drive continuous search improvement.

Implemented agile development principles to streamline processes and optimize productivity throughout all software development phases. Led successful product launches by coordinating multidisciplinary teams and ensuring timely delivery within budget constraints. Overseeing the hiring process to attract top talent, and implementing training and development programs to build a highly skilled engineering team. Scaled the system to process approximately 6,000 events per minute, enabling the delivery of 2 million user notifications daily.

Led the foundational engineering efforts for Moglix's Fintech division (Credlix.com), designing and implementing a scalable and highly available microservices architecture. Managed a team of 15+ engineers and QA professionals, driving the development of critical in-house systems (LOS, LMS, CRM, Customer Onboarding flow), ensuring their efficient integration and robust performance. Optimized resource allocation, effectively managing budgets and manpower for multiple simultaneous projects to achieve high-performing system delivery.

Designed and deployed a high-performance search and content-based recommendation system engine that enabled customers to discover hotels tailored to their unique requirements and preferences. Implemented key functionalities focused on effective user discovery and personalized search experiences within the platform.

Architected and maintained the core search infrastructure, initially built on Apache Solr, ensuring high availability and performance for personalized movie discovery. Spearheaded the strategic migration from Apache Solr to Elasticsearch, optimizing indexing, query performance, and scalability for handling large datasets and complex personalized recommendations. Developed and refined a proprietary "taste algorithm" that analyzes over 1000 elements of user preferences, noting likes and dislikes and making knowledge-based predictions. The system's understanding of user taste became more evolved and recommendations more precise as users contributed through an innovative rating mechanism.