JEET SARKAR

Achieved a Global Rank of 549 in the highly competitive TCS CodeVita Season 12, showcasing exceptional problem-solving and coding skills among over 100,000 participants.

• Secured a Global Rank of 549 out of 100,000+ participants worldwide in TCS CodeVita Season 12, demonstrating advanced algorithmic and programming capabilities.

Developed an AI-powered KYC Verification System leveraging computer vision and NLP to automate document validation and fraud detection, securing a Top 10 Finalist recognition.

• Developed an AI-powered KYC Verification System using computer vision and NLP to automate document validation and fraud detection.
• Built a secure pipeline integrating face matching, ID extraction, and anomaly detection for seamless digital onboarding processes.
• Recognized as a Top 10 Finalist among pan-India participants, validating innovative problem-solving and technical execution.

Architected and developed a backend-first agentic AI e-commerce platform leveraging LLMs and NLP for conversational product search and personalized recommendations at scale.

• Architected a backend-first agentic system utilizing LLMs and NLP pipelines to handle conversational product search, intent resolution, and contextual decision-making at scale.
• Built a stateful agent orchestration layer, enabling multi-turn voice-based shopping experiences with maintained session context and reliable backend actions across services.
• Developed scalable REST APIs for product discovery, recommendations, cart workflows, and AI-powered recipe planning, ensuring clear separation of concerns and low-latency responses.
• Implemented a semantic recommendation engine using behavioral signals and vector embeddings, significantly improving ranking relevance and personalized product retrieval.

Architected and developed an open-source LLM-powered agentic CLI assistant for autonomous file generation and multi-module project scaffolding.

• Architected a fully agentic CLI development assistant using open-source LLMs, enabling autonomous file generation and end-to-end application creation from natural language prompts.
• Implemented a robust chain-of-thought reasoning engine to analyze logs, detect root-cause failures, and self-correct code errors automatically.
• Achieved highly reliable iterative code generation, significantly improving development efficiency and reducing manual debugging efforts.

Developed an AI chatbot providing real-time trading and stock market news, utilizing a custom NLP model for sentiment analysis without external API reliance.

• Implemented real-time news data scraping from diverse articles, eliminating reliance on external APIs like Google News for financial insights.
• Trained a custom local NLP model to understand market sentiments, providing nuanced and accurate financial insights.
• Utilized Python, React.js, and TypeScript to build a robust and interactive chatbot interface, enhancing user engagement.

Developed a full-stack hotel listing website using the MERN stack, integrating secure authentication and dynamic UI for an enhanced user experience.

• Implemented secure authentication with Passport.js, enhancing user data protection and access control.
• Integrated Cloudinary for efficient image uploads, optimizing media management and performance for the platform.
• Designed a responsive and dynamic user interface using EJS templates, significantly improving user engagement and navigability.
• Ensured robust validation and role-based access control with Joi and custom middleware, bolstering system security and data integrity.

Gained practical experience in Deep Generative AI and implemented end-to-end data workflows from preprocessing to model deployment.

• Gained hands-on experience with advanced AI technologies, focusing on Generative AI and Deep Learning principles.
• Implemented end-to-end workflows, from data preprocessing to model deployment, ensuring practical application of theoretical knowledge.