Open to Software Engineer roles worldwide · Based in India

Hi, I'm Amardeep Bhagat

Software Engineer with 4+ years at ServiceNow, building production backend systems, REST APIs and event-driven services with Java & JavaScript. M.Tech in CSE from NIT Warangal.

Let's talk View my work
Amardeep Bhagat
4+ yrs @ ServiceNow
☕ Java · JS · Node
🎓 NIT Warangal
0Years experience
0Agents served by my system
0DSA problems solved
0JUnit tests added
01. about me

Engineer who owns it end-to-end

I'm a backend-focused Software Engineer who enjoys turning messy, real-world workflows into clean, reliable systems. For 4+ years at ServiceNow, I designed and shipped features used by hundreds of agents, from REST APIs and scheduling logic to automated workflows.

I care about the whole lifecycle: system design, implementation, testing and production support. I've cut flaky tests from 35% to under 5%, owned on-call for enterprise deliverables, and recently built MCP servers that let LLM agents use internal APIs as tools.

Outside work, I build full-stack cloud apps, solve DSA problems (1000+ and counting), and have authored 100+ verified CS answers as a Subject Matter Expert at Chegg.

  • 📍
    Based inHyderabad, India
  • 💼
    Most recentlySoftware Engineer (IC2), ServiceNow
  • 🎓
    EducationM.Tech CSE, NIT Warangal
  • ⚙️
    Core strengthsJava · REST APIs · Distributed Systems
  • 🚀
    Looking forSoftware Engineer roles, anywhere (open to relocation & remote)
02. experience

Where I've made an impact

ServiceNow

Jul 2022 — Aug 2026 · Hyderabad
Software Engineer (IC2)
Amardeep with his team at the ServiceNow Field Service Management demo showcase
📸 With my team at the ServiceNow Field Service Management (FSM) demo showcase
−30%data-inconsistency incidents
<5%flaky tests (from 35%)
−35%incident-triage time via MCP
+25%workflow throughput
  • Designed and shipped an end-to-end task-dependency management system with server-side validation and REST APIs used by web & mobile clients. It let 500+ agents manage dependencies through a conversational workflow, cutting data-inconsistency incidents by 30% and task-resolution time by 15%.
  • Built dependency-aware scheduling logic that validates task relationships and filters appointment slots by scheduling constraints and calendar views.
  • Improved dynamic resource allocation with worker-availability signals and fallback logic, reaching 95% automated-test coverage across scheduling workflows.
  • Added 200+ JUnit tests and fixed 90% of 120+ failing integration tests, bringing flaky tests down from 35% to under 5% and making CI/CD more reliable.
  • Automated a multi-step part-sourcing workflow in Java & JavaScript: 30% fewer manual steps, cycle time down from 4 days to 2.8 days, throughput up 25%.
  • Owned production on-call, resolving P2/P3 incidents with sub-2-hour MTTA, leading root-cause analysis and writing post-mortems.
  • Built custom Model Context Protocol (MCP) servers that expose internal APIs as structured tools for LLM agents, cutting incident-triage time by about 35%.
JavaJavaScriptREST APIsMicroservicesJUnitJenkins CI/CDRBAC & SecurityMCPEnterprise SaaS
03. projects

Things I've built

🌱

Heat Pump Bill Impact Tool

Full-stack clean-energy web app
  • React + TypeScript SPA with Express API and Azure Functions worker, modelling bill & CO₂ savings with client-side PDF reports.
  • Event-driven pipeline: API → Azure Service Bus → batch worker → MySQL/Prisma. Responds instantly and buffers writes under load.
  • Deployed on Azure through GitHub Actions with passwordless OIDC, Zod validation and Key Vault secrets.
ReactTypeScriptNodePrismaMySQLAzure
🤖

Multimodal Conversational Agents

M.Tech Thesis · NIT Warangal
  • Engineered a task-oriented dialog system that accepts both text and image input using a multimodal transformer architecture.
  • Reached a BLEU-4 of 0.4670, a 12% gain over the text-only baseline.
PythonPyTorchTransformers
🚗

Used Car Price Prediction

Machine learning · Flask web app
  • Predicts used-car resale prices from features like year, present price, km driven, fuel type and seller type, trained on CarDekho listings.
  • Compared Linear Regression with an Extra Trees Regressor, tuned with RandomizedSearchCV and 3-fold cross-validation.
  • Deployed the best model as a Flask web app for instant price estimates.
PythonScikit-LearnFlaskJupyter
View on GitHub →
04. skills

My toolbox

💻Languages

JavaJavaScriptTypeScriptC / C++Python

🧩Backend

REST APIsMicroservicesNode.jsExpress.jsEvent-Driven ArchitectureMessage Queues

🎨Frontend & Data

ReactHTMLCSSMySQLPrisma

☁️Cloud, DevOps & Testing

AzureGitHub ActionsJenkinsCI/CDJUnitAutomated Testing

🏗️Engineering

System DesignDistributed SystemsOOPDesign PatternsDSAGit

✨AI Tooling

MCP ServersClaudeCursorWindsurf
05. education & achievements

Learning never stops

2020 — 2022

National Institute of Technology, Warangal

M.Tech in Computer Science & Engineering · CGPA 7.2
Thesis: Multimodal Conversational AI Systems

2016 — 2020

Government Engineering College, Bilaspur

B.Tech in Computer Science & Engineering · CGPA 7.66

0

DSA problems solved across platforms

LeetCode↗ GeeksforGeeks ↗ HackerRank ↗ InterviewBit ↗

🏅 CSE Subject Matter Expert at Chegg Inc.: authored 100+ verified computer-science answers.

06. what's next?

Let's build something great together

I'm based in Hyderabad, India and open to Software Engineer roles anywhere: on-site, relocation or remote. Whether you have an opening, a referral or just want to connect, my inbox is open.

✉️ Say hello 📞 +91 83192 49879
✅ Email copied to clipboard