Suraj Kushwah - Software Architect

Executive Summary

I am a systems-first Software Architect who builds production-grade cloud systems, real-time media pipelines, and AI orchestration platforms. I combine deep Linux and infrastructure skills with product sense and people leadership. My core strength lies in turning ambiguous product requirements into pragmatic, observable, and secure systems that operate at scale under live pressure, such as cloud gaming, election monitoring, and multi-agent AI environments.

My strongest capability is listening to business requirements and instantly conceptualizing the entire system architecture. I can swiftly map out the necessary infrastructure from end-to-end, anticipate operational complexities, and design solutions that are resilient, high-quality, and deeply cost-optimized, focusing on giving us maximum control over the system.

"I prefer writing my own tools first, self-hosting second, and relying on third-party managed services as an absolute last resort."

I am actively seeking senior Software Architect roles at product companies, while openly maintaining my ongoing product work, advisory, and consulting engagements.

Work Experience

BugsWriter Company Pvt. Ltd. - Founder & Software Architect

India; 2023 - Present

Cyoni.ai (Multi-Agent AI Orchestration): Architected a multi-agent AI orchestration platform with a FastAPI Biz API, AiSys routing engine, MCPHub agent aggregation, Redis caching, PostgreSQL tenancy, and Svelte/Flutter clients. Led a cross-functional team of ~8 engineers.
AI Workflow & Compliance Automation: Built an AI-enabled compliance verification product for GDPR, ISO, and GRC validation via LLMs, along with an AI email classifier and form autofill browser extension based on context inference.
Internal Developer Pipelines: Embedded LLM tools (Gemini, Claude) into developer workflows to scaffold services, generate infrastructure templates, and automate refactoring and testing tasks.
Commercial & Operational Engineering: Designed business API surfaces (auth, billing), enforced per-tenant usage quotas, and deployed scalable microservices on GCP using API Gateways, IAM policies, and Terraform.

OnePlay Cloud Gaming - DevOps / Infrastructure Architect

Mumbai, India; July 2023 - April 2025

Cloud Gaming Infrastructure: Designed and operated multi-environment infrastructure across AWS and GCP, successfully leading a zero-downtime migration from AWS to GCP to optimize costs and system availability.
GPU & Data Center Operations: Managed GPU-equipped servers in colo (Yotta Data Center), optimizing session placement strategies and resource utilization to minimize idle GPU costs and meet strict low-latency SLAs.
Kubernetes & Microservices: Provisioned and managed Kubernetes clusters for microservices and non-GPU workloads, utilizing API Gateways for session lifecycle routing and microservice ingress.
Reliability, Security & Automation: Built comprehensive observability stacks using Prometheus, Grafana, and ELK; implemented automated failovers, deployed Vault, GuardDuty, and WAF; automated provisioning via Terraform and Ansible.

Bipolar Factory - DevOps & Backend Engineer

Bangalore, India; Nov 2021 - Apr 2023

AI Surveillance Pipelines: Built real-time CCTV inference systems utilizing NVIDIA DeepStream and TensorRT to accelerate video inference, capturing frames and alerting dynamically via distributed REST APIs.
Election Monitoring Systems: Managed 200+ RTSP streams for live election monitoring with custom dashboard overlays, deploying scalable infrastructure to handle high concurrency during voting hours.
Low-Latency Video Streaming: Designed WebRTC to RTMP conversion architectures using FFmpeg and SRS for distributed streaming, applying AI load prediction for dynamic stream distribution.

Key Skills & Technologies

Cloud & Infrastructure: Google Cloud Platform, Amazon Web Services, Linode, colo / Yotta (on-prem GPUs), VPC, API Gateway, IAM
Containers & Virtualization: Kubernetes, Docker, Helm, systemd, Proxmox, KVM
Backend & APIs: Python, FastAPI, Flask, REST, gRPC, WebSockets
AI & LLM Integration: LLM orchestration, Gemini, Claude, embeddings / RAG, prompt engineering, TensorFlow, PyTorch, inference pipelines
GPU & Acceleration: NVIDIA GPUs, DeepStream, TensorRT, GPU scheduling & session placement
Streaming & Media: FFmpeg, WebRTC, RTMP, RTSP, low-latency streaming patterns
Databases & Messaging: PostgreSQL, MySQL (Percona), MongoDB, Redis, Kafka, RabbitMQ
Infra Automation & CI/CD: Terraform, Ansible, Jenkins, GitLab CI/CD, IaC patterns, deployment pipelines
Observability & Ops: Prometheus, Grafana, ELK (Elasticsearch/Logstash/Kibana), Zabbix, synthetic tests, alerting, runbooks
Security & Secrets: Vault patterns, IAM best practices, WAF, GuardDuty, token/scoped API keys, multi-tenant isolation
Scripting & Tools: Bash, Python automation, FFmpeg pipelines, custom orchestration scripts
Frontend & Mobile: Svelte (web UI), Flutter (mobile client)
Product / Architecture Skills: Multi-tenant SaaS design, subscription & billing gating, quota enforcement, cost optimization, system design for reliability and operability

Notable Projects & Open Source

Explore my open-source work on GitHub.

CCTV Prompt Alerts: RTSP/HLS to FFmpeg frame extraction, Redis-queued to local GPU/cloud inference for LLM-based alert generation.
ffmagic: Asynchronous DVR system using Python and FFmpeg with multiprocessing for optimized archival and replay.
tuxi: Shell-based CLI assistant for Linux automation targeting power users.
offerclub: Full-stack eCommerce platform built from scratch with authentication and order management.
bugsnavi: OS/installer scripting project for automated Arch Linux deployment.

Leadership & Community

BugsWriter YouTube Channel: Grew a technical community of ~30,000 subscribers, producing content on Linux, self-hosting, and systems infrastructure.
Mentorship & Hiring: Recruited and mentored engineering teams, leading architectural reviews and technical onboarding.
Consulting & Advisory: Conduct active advisory sessions, aiding startups with architecture reviews, infrastructure optimization, and reliability improvements.