8+ years designing and operating multi-tenant cloud platforms across Kubernetes, AWS, and GCP. Built GenAI retrieval systems processing 5M+ events/dayand 500K+ daily querieswith 99.8–99.9% availability. Cut cloud spend 15–25%through predictive autoscaling and FinOps automation while reducing alert noise by 40–60%.
Okta • Lookout • CipherCloud • StartupManch | Bengaluru, India
Deep technical capabilities shaped by platform leadership, GenAI infrastructure, and cloud reliability at scale.
Led zero-downtime fleet upgrades across regulated production clusters, sustaining 99.8–99.9% availability. Established SLO/SLI governance, burn alerts, and incident review loops to harden reliability at scale.
Architected control/data planes for RAG pipelines processing 5M+ events/day and 100K–500K queries/day. Cut P95 indexing latency ~70% and maintained 99.8% SLOs under bursty AI traffic.
Built anomaly detection and signal correlation to reduce alert noise 40–60%, paired with ML-based capacity forecasting and autoscaling to cut cloud spend 15–25%+ without SLO regressions.
Quantified improvements across latency, cost, availability, and operational efficiency.
Reduced through async vector ingestion and event-driven architecture
Radical transparency on capabilities. What I own, what I use, and what I'm conquering next.
Battle-tested in production at scale. My primary weapons of choice.
Competent and capable. Used for enabling the core infrastructure.
Actively researching and rigorously working on to master.
From blockchain development to building planet-scale GenAI platforms—a journey of continuous growth
Contextualizing the pivotal shifts in infrastructure strategy across scale.
Co-Founder & Engineering Lead
Staff Software Engineer
Senior Software Engineer
Senior (Lead) Software Engineer
Contractor [Blockchain Consultant]
"Unpredictable 'burst' traffic from viral AI agents threatened to capsize the control plane, creating latency spikes > 5s."
Architected a 'Shared Control Plane' with predictive autoscaling and vector-caching layers. Decoupled ingestion from varying LLM latency profiles.
Full Whitepaper Restricted
Access requested. Document releasing soon.
Architecture: Event-Driven Vector Mesh
A living framework mapping technologies across adoption stages—from production-ready tools to emerging trends under evaluation
Container orchestration in production
Infrastructure as Code
Monitoring & alerting
GitOps deployment
Workflow orchestration
eBPF-based networking
Observability standard
Vector database for GenAI
Edge computing runtime
Developer portal
Cloud resource management
Distributed app runtime
Replaced with GitOps
Migrated to Helm 3
Replaced with K8s
Thought leadership distilled from building planet-scale systems. Real problems, real solutions.
Why your GenAI platform must be designed for worst-case latency, not averages. A deep dive into event-driven vector mesh architectures.
"Abstractions leak, but poor abstractions flood. In GenAI systems, you must build for the jagged edges of LLM latency, not just the averages."
At web scale, you cannot verify what you cannot observe. How to build observable systems that actually tell the truth.
"At web scale, hope is not a strategy. You cannot verify what you cannot observe. Observability is the only reality that matters."
Why automation isn't just about speed—it's about knowing the exact state of your infrastructure at any moment in time.
"Automation isn't just about speed; it's about the provenance of state. If a human has to touch production, the architecture has failed."
You cannot optimize what you haven't defined. How standardization unlocks velocity at scale.
"Standardization is the prerequisite to scale. You cannot optimize what you haven't defined."
Building immutable systems where trust is not a feeling but a mathematical certainty. Code is law.
"In Fintech, trust is an engineering specification, not a feeling. Code is law."
Monthly deep-dives on Kubernetes, GenAI infrastructure, and building platforms that scale. No fluff, just systems thinking.
💡 Connect newsletter service (Mailchimp, ConvertKit, etc.) to make this functional
Graduate-level training in software engineering and distributed systems, paired with telecommunications fundamentals.
BITS Pilani
Advanced Software Systems, Distributed Computing, Cloud Architecture
KIIT University
Computer Networks, Embedded Systems, Signal Processing
Contributing to the cloud-native ecosystem and sharing platform engineering insights.
Active contributor to cloud-native and platform engineering projects on GitHub.
View GitHub Profile→Sharing learnings on platform engineering, Kubernetes, and GenAI infrastructure through technical writing and community engagement.
Read Articles→Mentoring engineers toward senior/staff growth, running incident reviews, and leading architecture discussions across teams.
"Abstractions leak, but poor abstractions flood."
"At web scale, hope is not a strategy. Observability is the only reality that matters."
"Automation isn't just about speed; it's about the provenance of state."
"Standardization is the prerequisite to scale. You cannot optimize what you haven't defined."
"In Fintech, trust is an engineering specification, not a feeling. Code is law."
Building in public, sharing infrastructure learnings
I only work with 3-4 high-stakes organizations per year. Let's determine if we are a strategic fit.