Architecting Sovereign Intelligence & Distributed Agentic Workflows

I specialize in the design, deployment, and optimization of private AI ecosystems and fault-tolerant cloud architectures. As the Founder of Bhishaj Technologies (a registered MSME), my work bridges the gap between advanced machine learning and "Hard Engineering"—focusing on deterministic system reliability, strict ACID compliance, and measurable business ROI.

My engineering philosophy is rooted in Stoic-informed architecture: assuming system failure is inevitable and designing for absolute resilience through robust safety patterns, state-machine orchestration, and rigorous integration testing.

• 🌍 Based in Noida, UP
• ✉️ Contact me at gaurav.vashistha09@gmail.com
• 🏢 Founder of Bhishaj Technologies
• 🤝 Open to consulting on Sovereign AI, Agentic Workflows, and Cloud Architecture

AI & Multi-Agent Frameworks

Backend & Cloud "Hard Engineering"

Languages & Frontend

Media Processing & Specialized Systems

A private intelligence pipeline designed for strictly cited synthesis using native Indian LLMs.

• Architecture: Driven by the custom 'Sutra' pipeline, utilizing Sarvam 30B/105B models to retrieve and synthesize authenticated data exclusively from .gov.in sources. Integrated with multimodal STT/TTS capabilities.
• Business Outcome: Ensures absolute data sovereignty and semantic accuracy for regional-specific policy and governmental data queries, eliminating third-party data leaks.

A centralized operations hub deployed on GCP for automated business management.

• Architecture: Built on Google Cloud leveraging OpenClaw and Paperclip AI for agentic workflow orchestration. Employs robust state tracking and error handling.
• Business Outcome: Successfully automated 70% of freelance operations and project scoping, drastically reducing administrative overhead and accelerating client onboarding.

Distributed, fault-tolerant video bridging SaaS for seamless cinematic transitions.

• Architecture: Powered by Gemini 2.0 Flash and Veo 3.1 for generative transitions. Backed by Redis (implementing job safety via the BRPOPLPUSH reliable queue pattern) to guarantee task execution across distributed workers. Published a custom PyPI package (continuity-stitch) for highly optimized, frame-accurate 1080p normalization using FFmpeg.
• Business Outcome: Delivered a highly concurrent rendering pipeline that eliminates dropped frames and ensures zero-data-loss job processing, driving reliable high-tier SaaS subscriptions.

Autonomous multi-agent system for end-to-end e-commerce listing generation.

• Architecture: Orchestrates Gemini 2.5 Flash for high-fidelity vision analysis alongside Llama 3.3 (via Groq) for ultra-low-latency copywriting. Supported by a robust PostgreSQL database utilizing row-level locking to prevent race conditions during high-volume catalog ingestion.
• Business Outcome: Converts raw product images into high-conversion, SEO-optimized listings autonomously, maximizing merchant GMV velocity.

Project Delivery Engineer (AI Content & Automation Focus) Creative Lipi (IndiaMART Vendor) | Jan 2025 – Oct 2025

• Architectural Impact: Re-engineered the end-to-end catalog lifecycle by integrating advanced prompt engineering pipelines using Gemini and OpenAI.
• System Reliability: Designed deterministic data validation safety nets and implemented rigorous System Integration Testing (SIT), ensuring reliable processing at enterprise scale.
• Measurable ROI: Achieved a 40% reduction in manual overhead, successfully managing the autonomous validation and deployment of 200+ catalogs while maintaining 99% uptime.

• Qualcomm AI Upskilling Certificate: Technical Foundations (March 2026)
• Microsoft Learn Challenge: Create agents in Microsoft Copilot Studio (Jan 2025)
• Udemy: Flutter Development Bootcamp (Oct 2024)
• B.Tech in Computer Science & AI: ABESIT (2021 - 2024)
• Diploma in Computer Science: LPU (2018 - 2021)