Keshav Kant

As an AI Intern at Adda Education, I contributed to the development and deployment of AI-powered solutions for real-world educational use cases. My work involved designing and fine-tuning Large Language Models (LLMs) - Built n8n automation workflows that streamlined YouTube content management, paid-course delivery, and sales processes. - Designed and fine-tuned advanced text, audio models for personalized learning experiences and content analysis. - Developed GenAI-driven audit, insights, and reporting tools, boosting operational efficiency and scalability.

Providing 1:1 mentorship to students passionate about AI and web development. I specialize in guiding students through the process of building AI models from scratch, including training and testing deep learning models using TensorFlow and PyTorch. I also help students develop their own chatbots using Dialogflow and create full-stack web applications with the MERN stack. My mentorship focuses on hands-on projects and provides students with the skills and confidence to succeed in the field.

As an AI/ML Intern at Wellorgs Infotech, I contributed to the development and deployment of real-world AI solutions across education and productivity sectors. I engineered multimodal systems integrating voice, text, and vision-based models using GPT-4, ElevenLabs, and OCR pipelines. My work spanned full-stack integration, LLM tuning, and performance optimization. Key contributions: Developed a real-time voice-based AI assistant using GPT-4.1 and ElevenLabs API with sub-300ms latency via WebSockets Created an AI-powered exam evaluator capable of extracting and analyzing handwritten English and Arabic scripts using OCR and FAISS for content retrieval Fine-tuned LLM prompts and built RESTful APIs for scalable AI service integration across multiple apps (AI School, PEES, AI Counselor) Integrated Cloud Firestore and Flask for model hosting, user interaction logging, and content generation Improved latency by 6× and increased answer evaluation accuracy by 3× in live production demos This role sharpened my skills in NLP, multimodal AI, full-stack deployment, and building practical ML tools ready for real-world use.

As a Full Stack Developer intern, I led the development of both front-end and back-end solutions using React for dynamic user interfaces and Node/Express for robust server-side operations. Integrating essential APIs like Razorpay, Mailgun, and Firebase empowered me to create seamless, functional systems. This hands-on experience strengthened my skills in creating interconnected, efficient applications while navigating and solving complex technical challenges.

Completed an 8-week Salesforce Developer Virtual Internship program, where I gained expertise in Salesforce development by completing various Trailhead modules and Super Badges. I developed a comprehensive understanding of Salesforce fundamentals, organizational setup, process automation, and Apex development. I am proud to have showcased my proficiency in custom application development and automation of complex business processes.