┌─────────────────────────────────────────────────────────────────────────┐
│ │
│ ░█████╗░██╗░░░██╗██╗░░░██╗░██████╗██╗░░██╗ │
│ ██╔══██╗╚██╗░██╔╝██║░░░██║██╔════╝██║░░██║ │
│ ███████║░╚████╔╝░██║░░░██║╚█████╗░███████║ │
│ ██╔══██║░░╚██╔╝░░██║░░░██║░╚═══██╗██╔══██║ │
│ ██║░░██║░░░██║░░░╚██████╔╝██████╔╝██║░░██║ │
│ ╚═╝░░╚═╝░░░╚═╝░░░░╚═════╝░╚═════╝╚═╝░░╚═╝ │
│ │
│ M A N G A L · 0 x a y u s h M │
│ │
└─────────────────────────────────────────────────────────────────────────┘
const ayush = {
title: ["Founding Engineer @ Vinayak Industries", "Co-Founder @ BrewMyAgent"],
mission: "Build production systems where AI meets real businesses.",
currently: "Going deep on Context Engineering",
hardFacts: [
"Multi-tenant SaaS w/ Row-Level Security",
"Kafka pipelines at 250k+ IoT devices",
"RAG + LLM orchestration in production",
"AI that holds up on a Monday morning",
],
openTo: "Founders, builders & engineers working on something real → reach out",
};A multi-tenant industrial OS replacing spreadsheets, Tally & WhatsApp groups — across a 6-business conglomerate.
Manufacturing · Energy Storage · Fitness · Cables · Brushes · EdTech
| What | How |
|---|---|
| Real-time CEO dashboard | Next.js · Node.js · PostgreSQL (multi-tenant + RLS) |
| RAG-powered business insights | OpenAI · LangChain · Pinecone |
| Event-driven production tracking | Kafka · Docker |
| GST-compliant audit logs | PostgreSQL · Spring Boot |
| Legacy integrations | Tally · Payment gateways |
I own: Architecture · AI insight layer · Multi-tenant data model · All integrations
Custom AI agent studio shipping production automation for SMBs — lead management, content workflows, operational pipelines.
● TCS │ System Engineer │ Sept 2024 – Present
│ AI automation cutting financial workflow time by 50%
│ LLM-orchestration services in Python + Spring Boot
● Inventia │ Software Developer Intern │ Feb 2024 – Sept 2024
│ Event-driven Kafka pipeline → 250,000+ IoT devices in prod
│ DLQ fault-tolerance layer → zero data loss under peak load
Languages
Backend & Infrastructure
AI / GenAI
Frontend
Tooling
┌──────────────────────────────────────────────────────────┐
│ CONTEXT ENGINEERING │
│ │
│ The discipline of designing inputs around an LLM — │
│ system prompts, retrieved context, tool definitions, │
│ conversation history — so that production AI systems │
│ behave predictably at scale. │
│ │
│ Not vibe-coding. Not demos. Real trade-offs. │
└──────────────────────────────────────────────────────────┘
I write about what I actually learn building Vinayak OS and BrewMyAgent: multi-tenant data models · RAG evaluation · GPT-4 vs cheaper models · shipping AI to businesses that haven't seen it before.
→ Follow along on LinkedIn
I'm always open to conversations about:
[01] AI agents in production → what works, what breaks, what nobody warns you about
[02] Multi-tenant SaaS → especially in regulated / financial domains
[03] Event-driven systems at scale → Kafka in real customer environments, not tutorials
[04] Building for real businesses → the gap between AI demos & AI that holds up Monday morning
If you're a founder, builder, or engineer working on something interesting →
Building in public. Shipping for real customers. Learning out loud.