Software Architect

From protocols
to models.

20 years building systems at the boundary of infrastructure and intelligence — from network control planes to agentic AI. Based in Bengaluru, India.

View my work →
20+
Years of experience
20+
Patents accepted / filed
20+
Peer-reviewed publications
PhD
Computer Science, MSU
MIDS
UC Berkeley
India
Bengaluru · Remote
01

Experience

Sep 2025 – Present
Remote · Bengaluru, India
Member of Technical Staff
Startup
Applying agentic AI to the hardest problems in production systems reliability.
AI · Agentic AI · Observability · Reliability Engineering
Mar 2025 – Jul 2025
Bengaluru, India · Hybrid
Principal Engineer
Cisco
Engineering lead for the Cisco ACI controller team — owned observability, system recovery, and team execution across a distributed controller platform.
Golang · ACI · Distributed Systems · Observability · Team Leadership
Mar 2024 – Feb 2025
Remote
Engineer
Applied Intuition
Developed parallel, auto-scaled export workflows to enhance system performance and efficiency. Integrated CLIP-based multimodal models to enable natural language search and tagging over large-scale simulation data.
CLIP · Multimodal Search · API Development · Distributed Workflows · AI/ML
2021 – 2023
San Jose, CA
Architect
Zscaler
Data security and analytics platform architecture at scale.
Data Security · Cloud Architecture · Analytics · Zero Trust
2021
Remote / Austin, TX
Principal Software Engineer
FeatureBase (formerly Molecula)
Architecture and design of Molecula as a Service — control plane for multi-tenant feature store management, deployment orchestration, and analytical query serving.
Feature Store · Multi-tenancy · Control Plane · ML Infrastructure · SaaS
2017 – 2020
Palo Alto, CA
Principal Software Engineer
Cisco Tetration
Connector framework for network telemetry (NetFlow / IPFIX), cloud security posture management for AWS, and LDAP annotation framework for flow telemetry.
NetFlow · IPFIX · Cloud Security · AWS · LDAP · Network Telemetry
2011 – 2017
Santa Clara, CA
Principal Software Engineer
Citrix Systems
Designed StyleBooks — a declarative configuration language for NetScaler ADC. Built Cisco ACI/RISE integrations and an eventually-consistent distributed key-value store for session persistence across clustered nodes.
DSL · Compiler · Distributed KV Store · SDN · Cisco ACI · NetScaler
2006 – 2011
San Jose, CA
Senior Software Engineer
Cisco Systems
Control plane software for IP, MPLS, Traffic Engineering, and Fast Re-Route on the ASR9K edge router. Scale optimizations for routing protocol convergence.
IP · MPLS · Traffic Engineering · Fast Re-Route · Control Plane · ASR9K
02

Selected Projects

UC Berkeley · MIDS Capstone · 2023
Restor-AI-tion
An end-to-end AI pipeline to decipher Kuzushiji — ancient Japanese script readable by fewer than 0.01% of modern speakers. Combined deep learning OCR, K-Means clustering for reading order, and GPT-4 for correction and translation into English. UC Berkeley MIDS Capstone, 2023.
OCR · Unsupervised Learning · NLP · GPT-4 · Computer Vision
Project page ↗
UC Berkeley · MIDS · 2023
Signet-ring: Authenticating Sources and Lineages of Digital Objects
A cryptographic framework for establishing provenance and integrity of digital artifacts. Enables verifiable chains of custody across distributed systems — published at HCSS 2023.
Cryptography · Provenance · Digital Security · Hashing
PDF ↗
UC Berkeley · MIDS · 2022
Urban Land Use Classification from Public Domain Imagery
Deep learning classification of 21 urban land use categories using UC Merced aerial imagery. Applied transfer learning on pre-trained CNNs to achieve high accuracy on a publicly available remote sensing dataset.
CNN · Transfer Learning · Remote Sensing · Image Classification
PDF ↗
UC Berkeley · MIDS · 2022
Will Your Flights be Delayed?
Predicting 15-min+ flight delays using weather and airline operational data collected 2 hours before scheduled departure. Compared multiple models including XGBoost, logistic regression, and random forests.
Supervised Learning · Feature Engineering · Time Series · XGBoost
Report ↗ Slides ↗
UC Berkeley · MIDS · 2021
What Made the Pacific Northwest Move?
Statistical analysis of how vaccination rates influenced vacation mobility in the Pacific Northwest during summer 2021. Used causal inference methods to isolate vaccination effects from confounding mobility trends.
Causal Inference · Regression · Mobility Data · Statistics
Report ↗ Slides ↗
Citrix Systems · 2011–2017
StyleBooks: Declarative Configuration for Citrix ADC ↗
A compiler-backed DSL enabling simplified, portable NetScaler configuration across cloud architectures. Wrote the compiler that generates Python packages from StyleBooks source — used in production at Citrix.
Compiler · DSL · Declarative Programming · Network Automation
03

Selected Publications

2023
Signet-ring: A framework for authenticating sources and lineages of digital objects
HCSS · with C. Jimerson, D. Rorie, A. Samra, A. Mande, D. Aranki
Cryptography · Provenance · Digital Security · Hashing
PDF ↗ Paper ↗
2015
"Slow is Fast" for Wireless Sensor Networks in the Presence of Message Losses
JPDC · with R. Hajisheykhi, L. Zhu, M. Demirbas, S. Kulkarni
Distributed Systems · Fault Tolerance · Sensor Networks · Protocol Design
2009
Stabilizing Interference-Free Slot Assignment for Wireless Mesh Networks
Guide to Wireless Mesh Networks · with A. Jhumka, F. Abujard, S. Kulkarni
TDMA · Self-Stabilization · Wireless Mesh · Scheduling
2009
ProSe: A Programming Tool for Rapid Prototyping of Sensor Networks
S-Cube · with S. Kulkarni
Programming Tools · Sensor Networks · Code Generation · Rapid Prototyping
2004
A Line in the Sand: A Wireless Sensor Network for Target Detection, Classification, and Tracking
Computer Networks · Arora et al.
Sensor Networks · Target Tracking · Signal Processing · Distributed Detection
View all on Google Scholar →
04

About

I'm a software architect working at the intersection of AI and systems engineering — with a growing focus on agentic AI applied to production reliability and observability.

Over 20+ years I've built across the stack: routing protocols at Cisco, declarative configuration at Citrix, data security at Zscaler, and CLIP-based multimodal search at Applied Intuition. UC Berkeley (MIDS) sharpened the AI direction. After nearly 25 years in the US, I returned to Bengaluru in 2024 — building at the frontier of AI.

Education

MIDS
2023
UC Berkeley
PhD, Computer Science & Engineering
2006
Michigan State University
MS, Computer Science & Engineering
2003
Michigan State University
BE, Computer Science & Engineering
2001
College of Engineering, Guindy
05

Contact

GitHub LinkedIn Google Scholar