Spiritual AI — Personalized Guidance for How Your Mind Works

Professional Summary

Expert AI Engineer with 8+ years of theoretical and 4+ years of hands-on experience in Natural Language Understanding (NLU), Large Language Models (LLMs), and Retrieval-Augmented Generation (RAG). Specializing in designing and deploying scalable Conversational AI pipelines at production scale. Proven track record in model optimization (TensorRT, ONNX, Quantization) and multi-provider LLM orchestration. Dedicated to building intelligent systems that integrate advanced cognitive architectures with high-performance engineering.

Skills

AI/ML: LLMs (OpenAI, Gemini, Claude), RAG Pipelines, NLU, PyTorch, Hugging Face, FAISS, Qdrant, LlamaIndexOptimization & Deployment: TensorRT, ONNX, Quantization, Pruning, Model Serving (MLOps), Docker, CI/CDFull-Stack & Immersive: Next.js, React, Node.js, TypeScript, Three.js, WebGL, Python (FastAPI/Flask), SolidityInfrastructure: Arch Linux, Git, Vector Databases (pgVector), Supabase, Vercel, AWS/GCP Deployment

Projects

SpiritualAI.store — Conversational AI & RAG Platform2024 - Present

Architected and deployed a production-scale Conversational AI platform leveraging LLMs and a complex RAG pipeline using Qdrant and LlamaIndex.
Implemented multi-provider LLM orchestration with automated fallbacks and ensemble routing, increasing system reliability by 99.9%.
Optimized inference latency by 40% using TensorRT and Quantization (INT8), enabling real-time 3D interaction at 60 FPS on low-power devices.
Integrated semantic search and context management using FAISS, resulting in a 35% improvement in response relevance and a 25% increase in user retention.

Gesture AI — Vision-Driven NLU Interface2023 - 2024

Developed a high-performance computer vision system for real-time gesture-to-command translation using MediaPipe and custom PyTorch models.
Achieved 97.5% accuracy in complex gesture recognition through advanced data augmentation and transformer-based attention mechanisms.
Selected for IIT Delhi Startup Expo 2024 for pioneering work in non-tactile human-computer interaction and multimodal NLU.
Converted models to ONNX format for cross-platform deployment, reducing CPU overhead by 50%.

Cyber Security Agent — Automated Threat NLU2024

Engineered an AI-powered security monitoring suite using LLMs for automated log analysis and threat report generation.
Reduced false-positive threat detections by 22% through the implementation of a fine-tuned BERT model for anomaly classification.
Automated 90% of vulnerability reporting, saving 15+ hours of security engineering effort per week.

KalkiOS — AI-Native Operating System2024

Architected an AI-first desktop environment with integrated local LLMs for system-level automation and natural language shell interaction.
Developed a custom prompt engineering framework for system task execution, achieving 90% success in complex multi-step workflows.
Optimized local model inference using llama.cpp and hardware acceleration, enabling smooth performance on consumer hardware.

Notable Achievements

Exhibitor, IIT Delhi Startup Expo 2024: Recognized for innovation in Gesture-Based AI systems.
State Level Abacus Champion: Advanced cognitive foundation in logic and pattern recognition.
Open Source Leadership: Maintaining production-ready AI tools with significant community adoption on GitHub.

Education

Bachelor of Technology in Computer Science2022 - 2026

ABES Engineering College, Ghaziabad

Core Focus: Advanced AI, Data Structures, Distributed Systems, Operating Systems.

Professional Certificate in Product Management & Agentic AI2025 - 2026

IIT Patna (Masai)

Certifications

Data Science - KNIME (2024)

Microservices - Kong (2024)

Python Programming - OpenEDG (2024)

Advanced Pen Testing - LinkedIn (2024)

AI for Cybersecurity - PMI (2024)

C++ Programming - OpenEDG (2024)