Keerthana B V

KEERTHANA B V AI Agent & LLM Applications Developer +91 724479 __PROT 0__ Bengaluru Karnataka India __PROT 4__ __PROT 2__ Portfolio PROFESSIONAL SUMMARY Results driven AI Developer and MCA graduate with proven experience building secure full stack web applications and deploying autonomous LLM agents. Specializes in designing advanced RAG pipelines hybrid AI routing engines and custom automation workflows using Python Lang Chain and React. Passionate about bridging the gap between complex machine learning models and practical cost effective software solutions that drive real business value. TECHNICAL SKILLS Languages: Python __PROT 1__ (ES 6+) SQL AI & Machine Learning: Generative AI Large Language Models (LLMs) Retrieval Augmented Generation (RAG) Embeddings Lang Chain OpenAI API Claude API BERT NLP Chain of Thought Prompting Frontend & Backend: React__PROT 5__ Next__PROT 6__ REST APIs PostgreSQL MongoDB HTML 5 CSS 3 Dev Ops & Tools: Git __PROT 3__ Linux (Nginx) Vercel Netlify Methodologies: Agile Scrum SDLC Unit Testing Code Review Technical Documentation WORK EXPERIENCE Full Stack Developer ASPL Tech Solutions Pvt. Ltd. Bengaluru India Developed and maintained responsive web applications using React__PROT 7__ and Node__PROT 8__ successfully delivering client facing features for the healthcare and retail sectors on schedule. Directly collaborated with 5+ clients to gather technical requirements provide regular project updates and ensure the final software deliverables aligned with their business needs. Contributed to a custom HRMS platform by building an automated employee onboarding module which helped reduce administrative data entry time. Implemented secure authentication by integrating Role Based Access Control (RBAC) and JSON Web Tokens (JWT) for secure user logins and data protection. Assisted in application deployment by pushing code through CI/CD pipelines and helping host production sites on cloud environments. MERN Stack Developer Intern Dyashin Technosoft Pvt. Ltd. Bengaluru India November 2024 January 2025 Developed a full stack e commerce platform using MongoDB Express__PROT 9__ React__PROT 10__ and Node__PROT 11__ (MERN) featuring secure JWT based authentication and role management. Enhanced frontend performance using React Hooks and state management best practices improving page load speed by 25% and delivering a seamless cross device user experience. Designed and documented 15+ RESTful API endpoints with senior engineers for inventory tracking and order fulfillment following OpenAPI specifications. Resolved 15+ critical bugs during User Acceptance Testing (UAT) reducing the post launch defect rate by 40% and significantly improving platform reliability. KEY PROJECTS AI Powered Conversational Voice Bot & Google Sheets Integration Tech Stack: Vapi AI Twilio API Make__PROT 12__ OpenAI GPT 4 o Mini Google Sheets Webhooks JSON Schema. Generative AI Voice Agent: Deployed a multilingual conversational voice agent (English) using Vapi AI and OpenAI GPT 4 o Mini to automate customer order intake and L 1 FAQs. Google Sheets Integration: Built a serverless pipeline in Make__PROT 13__ using custom webhooks to automatically extract structured call data (name order details phone number) and log it directly as new rows in Google Sheets. Automated SMS Workflows: Integrated Twilio API to instantly send UPI payment links via SMS upon call ending adding a custom error handling route to prevent API delivery failures from stopping the order pipeline. AI Legal Document Intelligence Agent ( Link ) Stack: Legal BERT Py Torch Hugging Face OpenCV spa Cy Python React__PROT 14__ MongoDB Architected an autonomous AI Legal Agent utilizing fine tuned Legal BERT and Py Torch to automate clause extraction and introspective risk analysis achieving an 84__PROT 15__% F 1 score and 84__PROT 16__% accuracy on the 510 contract CUAD benchmark. Engineered an autonomous perception pipeline using OpenCV image binarization and adaptive thresholding to programmatically strip noisy stamp headers and footer margins boosting optical character recognition (OCR) text recovery by +32% on contaminated scans. Formulated a secure Trust & Safety guardrail agent featuring spa Cy powered PII redaction and a Human in the Loop feedback database (MongoDB) that captured specialist corrections to retrain models restricting algorithmic bias disparity gaps to under 10%. E Commerce Return Policy RAG Agent ( Live ) Stack: Python Lang Chain Llama 3__PROT 17__ (Groq API) Hugging Face Streamlit Built an AI Customer Support Agent using Python Lang Chain and the Groq API (Llama 3) to instantly answer user questions by securely scanning multiple company policy documents using local vector search (FAISS). Developed Enterprise CRM Integrations by dynamically injecting mock customer data (Names & Order IDs) into the LLM context allowing the AI to provide highly personalized support. Implemented an Automated Ticketing System featuring a smart routing protocol that instantly bypasses the AI to generate a support ticket and log chat history if an issue requires human escalation. EDUCATION Master of Computer Applications (MCA) August 2025 RV Institute of Technology and Management Bengaluru CGPA: 8__PROT 18__ / 10. Bachelor of Computer Applications (BCA) September 2023 Community Institute of Commerce and Management Bengaluru CGPA: 8__PROT 19__ / 10__PROT 20__ CERTIFICATIONS & ACHIEVEMENTS Google Cloud Professional: Prompt Engineering for Generative AI Google Cloud Skills Boost (2025) Google Cloud Credential: Introduction to Generative AI LLMs & Responsible AI Google Cloud Skills Boost (2025) MERN Stack Specialization: Industry Certification Dyashin Technosoft Pvt. Ltd. (2025) Best Paper Distinction: NCRIE 2025 National Conference on Recent Innovations in Engineering KSIT

KEERTHANA B V AI Agent & LLM Applications Developer +91 724479 __PROT 0__ Bengaluru Karnataka India __PROT 4__ __PROT 2__ Portfolio PROFESSIONAL SUMMARY Results driven AI Developer and MCA graduate with proven experience building secure full stack web applications and deploying autonomous LLM agents. Specializes in designing advanced RAG pipelines hybrid AI routing engines and custom automation workflows using Python Lang Chain and React. Passionate about bridging the gap between complex machine learning models and practical cost effective software solutions that drive real business value. TECHNICAL SKILLS Languages: Python __PROT 1__ (ES 6+) SQL AI & Machine Learning: Generative AI Large Language Models (LLMs) Retrieval Augmented Generation (RAG) Embeddings Lang Chain OpenAI API Claude API BERT NLP Chain of Thought Prompting Frontend & Backend: React__PROT 5__ Next__PROT 6__ REST APIs PostgreSQL MongoDB HTML 5 CSS 3 Dev Ops & Tools: Git __PROT 3__ Linux (Nginx) Vercel Netlify Methodologies: Agile Scrum SDLC Unit Testing Code Review Technical Documentation WORK EXPERIENCE Full Stack Developer ASPL Tech Solutions Pvt. Ltd. Bengaluru India Developed and maintained responsive web applications using React__PROT 7__ and Node__PROT 8__ successfully delivering client facing features for the healthcare and retail sectors on schedule. Directly collaborated with 5+ clients to gather technical requirements provide regular project updates and ensure the final software deliverables aligned with their business needs. Contributed to a custom HRMS platform by building an automated employee onboarding module which helped reduce administrative data entry time. Implemented secure authentication by integrating Role Based Access Control (RBAC) and JSON Web Tokens (JWT) for secure user logins and data protection. Assisted in application deployment by pushing code through CI/CD pipelines and helping host production sites on cloud environments. MERN Stack Developer Intern Dyashin Technosoft Pvt. Ltd. Bengaluru India November 2024 January 2025 Developed a full stack e commerce platform using MongoDB Express__PROT 9__ React__PROT 10__ and Node__PROT 11__ (MERN) featuring secure JWT based authentication and role management. Enhanced frontend performance using React Hooks and state management best practices improving page load speed by 25% and delivering a seamless cross device user experience. Designed and documented 15+ RESTful API endpoints with senior engineers for inventory tracking and order fulfillment following OpenAPI specifications. Resolved 15+ critical bugs during User Acceptance Testing (UAT) reducing the post launch defect rate by 40% and significantly improving platform reliability. KEY PROJECTS AI Powered Conversational Voice Bot & Google Sheets Integration Tech Stack: Vapi AI Twilio API Make__PROT 12__ OpenAI GPT 4 o Mini Google Sheets Webhooks JSON Schema. Generative AI Voice Agent: Deployed a multilingual conversational voice agent (English) using Vapi AI and OpenAI GPT 4 o Mini to automate customer order intake and L 1 FAQs. Google Sheets Integration: Built a serverless pipeline in Make__PROT 13__ using custom webhooks to automatically extract structured call data (name order details phone number) and log it directly as new rows in Google Sheets. Automated SMS Workflows: Integrated Twilio API to instantly send UPI payment links via SMS upon call ending adding a custom error handling route to prevent API delivery failures from stopping the order pipeline. AI Legal Document Intelligence Agent ( Link ) Stack: Legal BERT Py Torch Hugging Face OpenCV spa Cy Python React__PROT 14__ MongoDB Architected an autonomous AI Legal Agent utilizing fine tuned Legal BERT and Py Torch to automate clause extraction and introspective risk analysis achieving an 84__PROT 15__% F 1 score and 84__PROT 16__% accuracy on the 510 contract CUAD benchmark. Engineered an autonomous perception pipeline using OpenCV image binarization and adaptive thresholding to programmatically strip noisy stamp headers and footer margins boosting optical character recognition (OCR) text recovery by +32% on contaminated scans. Formulated a secure Trust & Safety guardrail agent featuring spa Cy powered PII redaction and a Human in the Loop feedback database (MongoDB) that captured specialist corrections to retrain models restricting algorithmic bias disparity gaps to under 10%. E Commerce Return Policy RAG Agent ( Live ) Stack: Python Lang Chain Llama 3__PROT 17__ (Groq API) Hugging Face Streamlit Built an AI Customer Support Agent using Python Lang Chain and the Groq API (Llama 3) to instantly answer user questions by securely scanning multiple company policy documents using local vector search (FAISS). Developed Enterprise CRM Integrations by dynamically injecting mock customer data (Names & Order IDs) into the LLM context allowing the AI to provide highly personalized support. Implemented an Automated Ticketing System featuring a smart routing protocol that instantly bypasses the AI to generate a support ticket and log chat history if an issue requires human escalation. EDUCATION Master of Computer Applications (MCA) August 2025 RV Institute of Technology and Management Bengaluru CGPA: 8__PROT 18__ / 10. Bachelor of Computer Applications (BCA) September 2023 Community Institute of Commerce and Management Bengaluru CGPA: 8__PROT 19__ / 10__PROT 20__ CERTIFICATIONS & ACHIEVEMENTS Google Cloud Professional: Prompt Engineering for Generative AI Google Cloud Skills Boost (2025) Google Cloud Credential: Introduction to Generative AI LLMs & Responsible AI Google Cloud Skills Boost (2025) MERN Stack Specialization: Industry Certification Dyashin Technosoft Pvt. Ltd. (2025) Best Paper Distinction: NCRIE 2025 National Conference on Recent Innovations in Engineering KSIT