The Frontier AI & Agentic Ecosystem
Resence is an independent AI technology laboratory engineering multi-model intelligence hubs, multimodal medical demystifiers, and autonomous Model Context Protocol (MCP) commerce protocols.
Products & Frontier AI Architectures
A cohesive suite of multi-model AI assistants, medical OCR intelligence, agentic commerce protocols, and social utility platforms under the resence.in domain umbrella.
Resence AI
Multi-Model AI Assistant & Intelligence Hub
Hardware-accelerated AI assistant orchestrating multiple frontier LLM models (Groq Llama 3.3 70B, Google Gemini 2.5 Flash, NVIDIA Nemotron 3.5, DeepSeek R1) with 100% prefix-locked KV-caching, multimodal vision, dynamic PPTX/PDF generator, and real-time live web search.
WellBridge AI
Patient Health Journal & Multimodal Medical Demystifier
Frontier healthcare companion developed for Google Cloud Gen AI Academy APAC (Cohort 3). Ingests complex multi-page pathology and radiology reports with Gemini Multimodal OCR and translates findings into clear patient guidance.
RazorAgent MCP
Model Context Protocol (MCP) Autonomous Commerce Engine
Published NPM package enabling AI reasoning models (Claude, ChatGPT, Hermes) to execute deterministic payments, link generation, and refunds with cryptographic SHA-256 idempotency locks. Built for Razorpay AI Buildathon 2026.
ServeMATE
Gamified Social Impact & NGO Donation Platform
Platform connecting individual donors and corporate CSRs with verified grassroots NGOs. Features Gemini AI donation advisor, real-time campaign transparency, and automated Razorpay checkout workflows.
Resence Fitness
AI Workout Architecture & Nutrition Tracker
Personalized fitness and biomechanics companion that formulates customized high-intensity split routines, calculates macronutrient targets, and tracks strength progression over time.
Anytime Converter
Real-Time Sign Language Interpreter & Document Suite
Computer vision assistive tool powered by PyTorch BiGRU and MediaPipe 126D hand landmarks for real-time sign language recognition, paired with high-performance client-side PDF manipulation utilities.
Built for Extreme Speed & Reliability
How the Resence technology ecosystem powers 6 interconnected subdomains with frontier models, deterministic protocols, and cloud infrastructure.
Multi-Model Neural Core
Unified orchestration across Groq Llama 3.3 70B, Google Gemini 2.5 Flash, NVIDIA Nemotron 3.5, and DeepSeek R1 with sub-second failovers.
450 t/s Hardware Acceleration
LPU-accelerated inference stream delivering instant responses with 100% deterministic prefix-locked KV-caching architecture.
Model Context Protocol (MCP)
Standardized agent rails enabling autonomous reasoning agents to safely trigger deterministic payments, refunds, and tools via NPM package v1.1.5.
Multimodal Vision & OCR
High-resolution document and medical pathology ingestion pipeline powered by Gemini 2.5 Flash with sub-millimeter bounding boxes.
Neural Speech & Audio Engine
Bi-directional audio pipeline combining low-latency Whisper speech-to-text with continuous Web Speech Synthesis across 99+ languages.
Serverless Cloud Infrastructure
Auto-scaling containers deployed across Google Cloud Run and Vercel edge nodes with global DNS resolution and zero idle overhead.
Led by Piyush Singh
Founder & Chief Executive Officer of Resence. Selected as a Google Student Ambassador (GSA '26, ID: 6403) and graduate of the Google Cloud Gen AI Academy. Directing autonomous agent systems to bridge frontier foundation models with high-utility applications.
Let's Build the Future of AI Together
Whether for collaboration, hackathons, AI architecture inquiries, or engineering partnerships, reach out directly.
Direct Email
Primary inbox for architecture inquiries and official correspondence.
th.piyushsingh2007@gmail.com
LinkedIn Network
Professional connections, leadership updates, and industry networking.
2,077+ Followers · Active Creator
GitHub Engineering
Public repositories, open-source MCP tools, and architecture codebases.
24 Repositories · @Piyush-Thakur7