Ashutosh Sharma

Mern Full Stack Developer at Worko.ai (2024-08 – Present)

Designed and implemented a Retrieval-Augmented Generation (RAG) job listing and search system leveraging AWS Lambda, embeddings, and OpenSearch vector search to enable semantic retrieval and context-aware job recommendations beyond traditional keyword matching.
Enhanced AI-driven resume-job matching accuracy by 30% through integrating large language model (LLM)-based semantic matching with structured rule-based scoring and continuous recruiter feedback loops.
Developed a scalable score generation wrapper to match resumes with job descriptions, supporting bulk parsing of up to 1,000 resumes per user using Azure AI for document extraction and AWS SQS for queue-based batch processing, incorporating isolated retry logic for reliable large-scale processing.
Collaborated in developing an AI-powered video interview platform by integrating backend AI services using Blazeface and TensorFlow, enabling automated candidate evaluation and insightful analytics.
Led the migration from Angular to a robust React-based architecture, implementing secure storage wrappers, custom API and toast handlers, global error boundaries, and reusable UI components.
Utilized React Hook Form, React Query, and Context API to optimize state and data management, and integrated AWS Amplify for centralized authentication workflows.
Developed a shared microfrontend component library to standardize UI/UX, accelerate cross-team development, reduce code duplication, and promote reusability.

Full Stack Web Developer at My Regular Trip (2023-09 – 2024-07)

Optimized MongoDB integration for user authentication, enhancing data retrieval speed by 20%.
Developed secure, robust middleware for user authentication, reducing authentication errors by 25%.
Spearheaded end-to-end development of a React application with 100+ dynamic screens, boosting user engagement by 40%.
Integrated Razorpay for seamless payment processing and real-time synchronization, cutting transaction response time by 35%.
Implemented OAuth 2.0 protocol, improving user satisfaction and retention rates by 50%.
Built an automated email marketing system using Mailjet to dispatch time-targeted promotional campaigns, increasing repeat bookings.
Designed and stored reusable dynamic email templates in MongoDB to enable personalized content rendering during sends.
Implemented Redis for efficient refresh token storage and session management, facilitating rapid token lookup and revocation.
Added IP-based rate limiting with Express Rate Limiter, automatically flagging and blacklisting abusive IPs in Redis to safeguard API endpoints against brute-force and DDoS attacks.
Developed a real-time customer support ticketing system that captures user queries and displays them on an admin dashboard.

Software Engineer at MSE Trade Promotion Council (2023-03 – 2023-08)

Built a responsive website for displaying user stats and stock details (e.g., stock quantity, future stocks), increasing engagement by 25% through a seamless UI.
Implemented FIX API to enhance algorithmic trading experience using the QUICKFIX engine.
Developed REST APIs with Express.js, integrated JWT authentication, and followed the MVC pattern, reducing query latency and API response time by 20%.

About