Open to opportunities

Shubh Rai

@shubhrai

Message

Software Engineer building AI agents for secure, scalable platforms.

India

Message

What I'm looking for

I want to build AI-powered systems that ship real security and data workflows—agent engines, integrations, and scalable pipelines—inside product-minded teams. I’m excited by hardening for reliability (RBAC, audit logs, air-gapped readiness) and fast iteration.

I’m a Software Engineer (SDE-1) at ZySec AI, where I’ve built ZySec’s flagship Multi-Tenant AI SOC/SOAR platform end-to-end—from alert ingestion and investigations to SOAR playbooks with approval gates and compliance reporting. I focus on turning SOC and leadership requirements into shipped product quickly, using AI-assisted development tools.

Beyond the platform, I built an AI Task & Agent Engine supporting 43 task types with Dockerized execution modes, approval gates, heartbeats, and per-run decision logs. I also shipped 105 integrations across SIEM/EDR/ticketing/IAM/threat intel/AML use cases with credential-rotation and a JWT bridge, implemented Casbin RBAC with secure, tamper-evident audit logging, and engineered a distributed ClickHouse/Dask/Prefect data engine processing 1B+ records—reducing query latency from minutes to milliseconds. I’ve also worked as an AI Data Engineer (freelance) delivering LLM fine-tuning datasets and AWS-backed ETL pipelines, and I build product-grade prototypes like an open-source AI memory plugin and a full-stack project-management SaaS + CLI.

Experience

Work history, roles, and key accomplishments

Current

SDE - 1

Current

ZySec AI

Jun 2025 - Present (1 year 1 month)

Designed and built ZySec’s multi-tenant AI SOC/SOAR platform end-to-end, including alert ingestion and triage, investigations and case management, and SOAR playbooks with approval gates. Built an AI task/agent execution engine with integration ecosystem and implemented RBAC, tamper-evident audit logs, and hardened deployments for air-gapped environments.

Python fastAPI MongoDB Vue.Js TypeScript Clickhouse LangGraph Docker RBAC Casbin JWT MFA Audit Logging Playwright Distributed Data Processing Dask Prefect E2E Testing

AI Data Engineer

MavnAI

Jul 2024 - Aug 2025 (1 year 1 month)

Delivered domain-specific training and evaluation datasets for LLM fine-tuning to improve model accuracy. Automated ETL pipelines using Python and SQL with AWS-backed storage workflows to reduce manual preparation time.

Python SQL AWS BASH ETL Data Pipelines LLM Fine tuning