Work · Full detail · status: online

Work

Staff Software Engineer with 16+ years across the stack: agent architectures, tool orchestration, and evals first, backed by deep React/TypeScript frontend craft and hands-on Python/Java backend ownership. Founded ServiceNow's Agentic Developer Platform (adopted by 20+ teams via hands-on onboarding), led a 9-person engineering team there, and designs agent systems end-to-end — tool contracts and orchestration, eval-gated rollout (golden sets, LLM-as-judge), governance via AI Agent Studio, shipped ALA from requirements to GA. Ex-co-founder with direct client delivery experience; enterprise-grade UI with testing and accessibility.

Download CV (PDF) → Read the writeups → Side projects →

What I do

Agentic AI & MCP

Designed and shipped MCP-driven agentic workflows integrated into the engineering lifecycle, adopted by 20+ teams with roughly 95% less research and debugging effort.

MCP, Claude Code, Build Agent, Tool orchestration, Self-healing loops

Local LLM inference

Run MLX-optimized models (Qwen, Gemma, DeepSeek) locally on Apple silicon via Ollama, plus memory-efficient loaders like turbo-fieldfare that stream model portions on demand: Gemma 4 26B under 2GB RAM. Took VLMs end-to-end on-device: Qwen2-VL-2B on llama.cpp/GGUF via Metal across a quantization ladder on iPhone, with sub-second TTFT from architecture and resolution levers.

Ollama, MLX, turbo-fieldfare, Apple silicon, Memory-efficient inference, llama.cpp, GGUF quantization, On-device inference

Developer platforms

Founded the Agentic Developer Platform: multi-team bootstrap and governance infrastructure for agents, skills, and MCP servers.

Multi-team, Governance, Telemetry

Frontend architecture

Architected scalable React and TypeScript systems powering low-code tools like Table Builder and Workspace Builder.

React, TypeScript, Low-code

Backend & data ownership

Hands-on backend behind the surfaces: Python/FastAPI and SQLite services with guarded, read-only SQL access, plus Java/Glide enterprise case backends and REST APIs.

Python, FastAPI, SQLite, Java / Glide, REST APIs

System design

Metadata-aware, multi-model architectures and design patterns built to hold up across teams, tenants, and releases.

Scalable systems, Design patterns

Quality & reliability

Enterprise testing strategy with Jest, React Testing Library, and Playwright; performance optimization and WCAG/ARIA accessibility. Lifted legacy coverage from 20% to 90%.

Jest, RTL, Playwright, WCAG/ARIA, Performance

Technical leadership

Lead a 9-person engineering team, run technical hiring and high-bar architectural reviews, and mentor engineers in AI-first development practices.

Hiring, Mentoring, Reviews

Agentic Developer Platform

Multi-agent orchestration. 20+ skills, 20+ agents, 20+ teams, four weeks from concept to production.

ADP coordinator multi-agent pattern MCP serverstools and context skills, 20 plusbootstrap agents, 20 plusorchestrated governancetelemetry
One session bootstraps a squad: skills, agents, rules, MCP servers.

Selected work

run 01 · ServiceNow✓ shipped

Agentic Developer Platform (ADP)

Creator, ServiceNow

Multi-team bootstrap and governance infrastructure for Claude Code, MCP servers, agents, and skills. Built in four weeks on a multi-agent orchestration pattern, rolled out to 7 pilot squads across 20+ teams with 20+ skills and 20+ agents.

✓ −90% task resolution time
✓ 20% → 90% legacy coverage
✓ open defects → single digits org-wide
✓ Days → hours
✓ Multi-agent pattern
Read the deep dive →
run 02 · ServiceNow✓ shipped

MCP-driven engineering workflows

20+ teams, ServiceNow

Integrated MCP-driven agentic workflows into the engineering lifecycle organization-wide. Reduced research and debugging effort by approximately 95% while improving accuracy.

✓ −95% research & debugging
✓ 20+ teams
✓ Higher accuracy
Read the deep dive →
run 03 · ServiceNow✓ shipped

Build Agent, ServiceNow Studio

App Engine, ServiceNow

Elevated App Engine with agentic workflows for AI-assisted application creation, flow generation, and tool orchestration across 35+ metadata types and 11 domains. Built on the planning pattern (think first, then execute) with a metadata-aware, multi-model (Claude default; Now LLM + Claude on Bedrock), self-healing agentic architecture.

✓ 35+ metadata types · 11 domains
✓ Multi-model
✓ Self-healing
✓ Metadata-aware
✓ Planning pattern (think → execute)
Read the deep dive →
run 04 · ServiceNow✓ shipped

Release lifecycle & app summary agents

Now Assist for Creator, ServiceNow

Automated delta analysis, human-readable release notes, and CAB-ready deployment manifests: documentation time from hours to a single click. Built on the REAct agentic framework (reason → act → observe). Ships as the Release lifecycle documentation agent in Now Assist for Creator.

✓ Hours → one click
✓ CAB-ready manifests
✓ Delta analysis
✓ REAct framework
Read the deep dive →
More writeups: SpendIQ Copilot: guardrails, evals, and honesty for an agent over enterprise data, On-device VLM on iPhone: 270 benchmark runs on the quantization ladder, Local LLM inference on Apple silicon: Ollama, MLX & turbo-fieldfare, Claude Code rollout to 20+ teams in 4 weeks, Table Builder, Workspace Builder, Revision: a local-first FSRS-5 desktop app with drag and air gestures.
All posts →
Telemetry

Results

Every number below is sourced from the work it describes.

95%
less research and debugging effort for 20+ teams on MCP workflows (measured across the 7 pilot squads in the 4-week rollout)
−90%
case task resolution time on the Agentic Developer Platform
20% → 90%
legacy test coverage lift on ADP pilot squads
single digits
open external defects org-wide, down ~90%
80% → 95%
accessibility compliance lift, Angular-to-React migration at Reputation
9
engineers led and mentored as tech lead at ServiceNow
4 wks
from concept to production for a multi-team AI platform
80%
faster test-to-production HRMS setup, Oracle Benefit Plan Copy
30s → 805ms
TTFT for on-device VLM inference via architecture and resolution levers (Ordo)
Lab

On-device, on-desk

# desk: Apple silicon
$ ollama run qwen3 && turbo-fieldfare load gemma-4-26B
✓ 26B under 2GB RAM (streamed on demand)
# phone: iPhone A16, Qwen2-VL-2B, GGUF ladder, 270 runs
$ ordo bench --ladder q4_k_m..q2_k --device a16
✓ TTFT 30s to 805ms (architecture and resolution levers)
Writeups: On-device VLM on iPhone · Local LLM inference on Apple silicon

Experience

ServiceNow

Staff Software Engineer
Hyderabad, Oct 2023 to Aug 2026
  • Designed ALA's agent architecture (tool contracts, delta-analysis flow, eval checks); built requirements to delivery; EA Mar 2026, GA Apr 2026; one-click release notes and update-set descriptions, CAB-ready.
  • Set the consolidation architecture: App Summary from Now Assist skills to AI Agent Studio; one generic flow replacing 4-provider settings (Now LLM, Gemini, OpenAI, Anthropic) — less maintenance, smaller monthly model-change test surface.
  • Legacy test coverage for update sets + source control + team dev: <20% to 95%, Feb–Mar 2026 in 4 weeks, using ADP; demoed to 7 teams, 20+ teams adopted.
  • Stabilized inherited Table Builder (UI/API fixes); grew entry points 1→4: standalone homepage, legacy table/form-builder entry, App Engine, ServiceNow Studio.
  • Same for Workspace Builder: inherited, mostly built; owned Studio support and UI/API fixes.
  • Owned front-end performance for the builder surfaces: profiled and diagnosed production regressions, fixed render and bundle hotspots, held performance budgets.
  • Instrumented agent quality in production: reasoning traces, tool-selection and hallucination error tracking, rollback paths, cost/latency dashboards feeding failures back into evals.
  • Human-gated autonomy: chat → plan → explicit approval → execution → verification, with per-action approvals (script, intent, target scope) and rollback/hand-back paths — checkpoints structural, not advisory.
  • On-call and incident command for enterprise customer-facing surfaces.
  • Tech-lead 9-dev team on 2-week sprints (~8 team-wide customer cases/sprint); day-to-day ~60% hands-on coding, PR and architecture reviews, 10% mentoring, 30% cross-team coordination; full-stack case ownership across React surfaces and Java/Glide backend (update sets, source control, team dev, ATF); shadowed juniors on rotation.
AI Agents, MCP, Python, TypeScript, React, Node.js, Java

Rippling

Senior Software Engineer
Bengaluru, Oct 2022 to Aug 2023
  • Contributed to frontend infrastructure and global engineering initiatives; identified codebase problem areas, drove improvement proposals, interviewed candidates, and mentored junior engineers.
JavaScript, React, TypeScript, Node.js, Python, Playwright, Sentry, LogRocket

Reputation

Lead Engineer
Hyderabad, Sep 2020 to Sep 2022
  • Led the Actions Team frontend, migrating the application from Angular to React and cutting page load time in half.
  • Delivered the Actions feature and management dashboard; raised accessibility compliance from 80% to 95% for a broad user base.
  • Built maintainable, scalable, and testable features; mentored junior developers and reviewed code.
JavaScript, Node.js, React, Python, RTK Query, Sumo Logic

Worktech Software

Co-Founder / Lead Developer
Hyderabad, Jul 2017 to Aug 2020
  • Built products and services using JavaScript, Node.js, MongoDB, Meteor, and REST APIs.
  • flexzone.in (2018-2020): aggregator of cafés into coworking spaces.
  • hospeto.in (2017-2018): video consultation platform for doctors, with health record and billing management.
JavaScript, Node.js, MongoDB, Meteor, REST APIs

Eze Software Group

Senior Developer
Hyderabad, Jun 2016 to Jun 2017
  • Developed and improved an order-management system for real-time trading related to short selling.
  • Maintained EzeLocate and DataFeed for locating short inventory and real-time stock-data processing.
REST APIs, MS SQL

Oracle

Project Leader
Hyderabad, Jun 2010 to May 2016
  • Developed and improved Plan Copy and Benefits Relationships systems using Java, Oracle Database, Oracle ODI, and SQL/PLSQL.
  • Completed the Benefit Plan Copy project, reducing test-to-production HRMS data setup time by 80%.
Java, Oracle Database, Oracle ODI, SQL/PLSQL

Skills

AI & Agentic Engineering

AI agents, Agent UIs, MCP workflows, Claude Code, LLM APIs (Anthropic, OpenAI, Gemini), LangChain / LangGraph, Prompt engineering, Tool orchestration, Multi-model systems, Multi-agent orchestration, REAct agents, LLM evals (golden sets, LLM-as-judge), Agent evaluation & guardrails, Human-in-the-loop agent design (plan → approve → execute → verify), AI governance (AI Agent Studio / Control Tower), RAG / keyword retrieval (BM25, FTS5)

Frontend & Platform

JavaScript (ES6+), TypeScript, React, Next.js, Redux Toolkit, RTK Query, Node.js, Webpack, Vite, PWA

Backend & Data

Python, FastAPI, SQLite, Java, REST APIs, MongoDB

Architecture

System design, Scalable frontend architecture, Design patterns, Developer productivity, Metadata-driven development

Reliability & Quality

Jest, React Testing Library, Playwright, Performance optimization, Performance profiling & tuning (production regressions, budgets), WCAG / ARIA accessibility

CI/CD & Tooling

GitHub Actions, Jenkins, Docker

Observability

Sentry, LogRocket, Datadog, Sumo Logic, Agent observability (reasoning traces, tool-selection errors, cost/latency), On-call & incident command

Education

National Institute of Technology (NIT), Tiruchirappalli

Master of Computer Applications (MCA), 2007-2010

Banaras Hindu University (BHU), Varanasi

BSc (Hons), Mathematics, 2003-2006

Patent US20150046358

Importing tested objects into benefits programs deployed on production systems. Assigned to Oracle Corp., status Abandoned. Google Patents