JEET SARKAR

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 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 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 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.

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.

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 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.