Harish Duddupudi

000

Decoding 89 frames so the scroll feels expensive.

01 / About

Software Architect · Backend & Cloud-Native Platforms

Open to Principal Engineer and Software Architect roles.

I'm an AWS Certified Solutions Architect with more than 15 years spent defining and delivering scalable, secure, cloud-native platforms — across utility software, industrial IoT, and enterprise systems. I began in the energy sector, building distribution management, outage management, and SCADA-integrated grid automation tools for electricity utilities — the kind of system where a bad deploy is visible from the street. From there I moved into large-scale IoT device orchestration, with real-time telemetry and diagnostics pipelines running at fleet scale.

Today the work splits two ways. One half is platform architecture: service boundaries and data contracts, API governance, secure SDLC and DevSecOps, edge-cloud deployment. The other is agentic systems — agents, harnesses and reusable skills on Claude and GitHub Copilot that automate the parts of delivery people should not be doing by hand.

Both halves come back to the same thing: design reviews, pairing and cross-team workshops with the engineers who build on what I define.

Based in
Ontario, Canada
Experience
15+ years
Focus
Platform architecture, API governance, DevSecOps
Certification
AWS Certified Solutions Architect

0+

Years engineering platforms

Since 2011

0

Industry domains

Utilities · IoT · Life sciences · Banking

0+

Microservices delivered

Robert Bosch

0%

Infrastructure cost reduction

Robert Bosch

02 / Agentic Engineering

I automate the work, not just the deploys.

Alongside platform architecture I build agentic systems — agents, harnesses and reusable skills on Claude and GitHub Copilot — that take repetitive engineering work off people and hand it to something that can run unattended.

The current one models an entire Product Development Life Cycle as a set of cooperating agents. Every role in product development is its own agent, composed from a persona, a rule set and a library of skills, and each hands structured output to the next stage rather than to a human inbox.

Claude
Agents, harnesses and skill libraries — multi-agent orchestration with personas and rule sets.
GitHub Copilot
Agents and reusable components wired into day-to-day development workflows.
U.S. Bank
Reusable skills and components authored for the delivery pipeline — see the ledger below.

Harnesses in the delivery pipeline

U.S. Bank

At U.S. Bank the same approach runs inside a regulated delivery pipeline. Each harness is scoped to a surface the platform already governs, so the output lands as a reviewable artefact rather than as a suggestion nobody owns.

  • API threat-model reviewer

    Secure SDLC · API design

    Walks a new or changed API surface and drafts the threat model against it before the design review, so the review argues about findings instead of producing them.

    Claude

  • Contract & boundary check

    Service boundaries · data contracts

    Reads a proposed change against the published contract and flags breaking changes and boundary violations while they are still cheap to undo.

    Claude

  • Quality-gate triage

    SonarQube in CI/CD

    Separates real defects from noise in static-analysis output and attaches a suggested fix per finding, so a red gate is actionable rather than ignored.

    GitHub Copilot

  • API governance assistant

    APIGEE lifecycle

    Checks a proposal against the governance standard — routing, versioning, rate limiting, auth — and reports where it diverges.

    Claude

  • Release-readiness reviewer

    Environment promotion

    Assembles the evidence a promotion needs in one place: tests, gates, tracing coverage and a rollback path.

    Claude + Copilot

  • Design-review companion

    Architecture review · mentoring

    Prepares the questions a design review should ask from the architecture standards, so a junior engineer walks in with the same checklist a principal would.

    Claude

The Product Development Life Cycle below is where this goes next: the same idea, but the roles themselves become the agents.

Product Development Life Cycle · agent per role

  1. Discover
  2. Define
  3. Build
  4. Verify
  5. Release

Frames the problem before anyone proposes a solution: who it is for, what it is worth, and what happens if it is not built.

Skills composed

  1. Opportunity framing
  2. Market and competitor scan
  3. Outcome definition
  4. Roadmap sequencing
  5. Stakeholder synthesis

Hands off

  • Problem statement
  • Success measures
  • Prioritised opportunity list

Composed from persona + rules + skills

Arrow keys move between roles. Six of them, one pipeline, and not one has ever asked to move the standup.

03 / Domains

Four industries, each with its own failure modes. Open one for the role, the case study and what it returned.

  • Banking & Payments

    2023 – Present

    U.S. Bank

    Merchant lending platform across microservices processing millions of daily transactions, under enterprise uptime SLAs and API governance.

    Detail
  • Life Sciences

    2021 – 2023

    GSK

    Architecture authority for a global enterprise platform: monolith decomposition, centralized identity, design standards across distributed teams.

    Detail
  • Automotive IoT

    2016 – 2021

    Robert Bosch

    Connected-vehicle device orchestration: real-time telemetry ingestion, diagnostic event processing, fleet analytics at scale.

    Detail
  • Energy & Utilities

    2011 – 2015

    Fluentgrid

    Grid automation for electricity distribution: distribution management, outage management, SCADA-integrated engineering tooling.

    Detail

04 / Approach

How the work gets done.

  1. 01

    Define the boundaries

    Service boundaries, data contracts and integration patterns settled before code — so the system has a shape teams can build against.

    U.S. Bank · GSK

  2. 02

    Govern the surface

    API lifecycle governance end to end: routing, versioning, rate limiting and security managed as a product, not an afterthought.

    APIGEE · U.S. Bank

  3. 03

    Secure the pipeline

    Threat-modelled API surfaces, static analysis gates in CI/CD, and centralized OAuth2 identity — security enforced by the pipeline rather than by review.

    SonarQube · OAuth2 · Azure AD

  4. 04

    Ship with parity

    Containerized rolling deployments across hybrid On-Prem and cloud, with full environment parity so what passes UAT is what runs.

    Rancher/Kubernetes · Azure

  5. 05

    Prove the failure paths

    Distributed tracing, structured logging, circuit breakers and chaos-tested recovery — plus FMEA across platform components.

    U.S. Bank · Robert Bosch

  6. 06

    Leave the team stronger

    Design reviews, pairing and cross-team workshops, so the architecture outlives whoever wrote the first version of it.

    Every role since 2016

Harish Duddupudi, Software Architect · Backend & Cloud-Native Platforms

Ontario, Canada

Fifteen years of shipping systems other people depend on.

05 / Experience

Four domains, one discipline.

Each chapter below animates what the platform actually did — payments crossing a gateway, a monolith decomposing, a vehicle fleet reporting in, a grid recovering from an outage.

Yes, those are real animations rather than screenshots. No, it was not the fastest way to write a CV.

Ch. 01

U.S. Bank

Lead Software Development Engineer

Feb 2023 – Present · Remote · Toronto, ON

Transactions streaming an API gateway perimeter: threats refused at the boundary, a tripped circuit breaker rerouting around failure, governance rings tightening on the core.

  • Defined end-to-end platform architecture for a merchant lending system: service boundaries, data contracts, API governance, integration patterns across microservices processing millions of daily transactions.
  • Instituted a secure SDLC: threat-modelled API surfaces, SonarQube in CI/CD, OAuth2 identity governance, APIGEE rate-limiting and security.
  • Led API lifecycle governance via APIGEE — routing, versioning, security.
  • Deployed containerized services across On-Prem + Azure hybrid using Rancher/Kubernetes with rolling updates and full environment parity.
  • Added distributed tracing, structured logging, circuit breakers, and chaos-tested recovery to meet enterprise uptime SLAs.
  • Mentored engineers through design reviews, pairing, and cross-team workshops.
  • Java 17
  • Spring Boot
  • Kafka
  • PostgreSQL
  • OAuth2
  • APIGEE
  • GraphQL
  • Rancher/Kubernetes
  • Azure
  • Autosys
  • Docker
  • GitLab CI/CD
Ch. 02

GSK

Software Architecture Lead

Dec 2021 – Feb 2023 · Remote · Toronto, ON

A monolith fracturing into independently deployable services: dependency chords thinning as they decouple, a centralized identity ring closing around the cluster.

  • Architecture authority across a global enterprise platform: blueprints, technology evaluation, design standards across distributed teams.
  • Migrated legacy Struts/WebLogic monoliths to Spring Boot microservices — 60% reduction in inter-module dependencies, independent deployability.
  • Centralized authN/authZ with OAuth2 + Spring Security + Azure AD.
  • Built a microservice scaffolding generator — 40% faster provisioning of new services, with platform consistency enforced by template.
  • 50% improvement in transactional and reporting performance via Oracle PL/SQL optimization, indexing, and data-access redesign.
  • Jenkins/Maven CI/CD with embedded SonarQube quality gates across Dev/UAT/Prod, eliminating manual promotion errors.
  • Java
  • Spring Boot
  • Spring Security
  • OAuth2
  • Oracle PL/SQL
  • Azure AD
  • AWS
  • Jenkins
  • Maven
  • SonarQube
  • Docker
  • Kubernetes
Ch. 03

Robert Bosch

Software Development Engineer III, IoT & Connected Systems

Nov 2016 – Nov 2021 · Bengaluru, India

A vehicle fleet on a receding road plane, each unit beaming telemetry to an ingest cloud, a diagnostic fault flaring, and the stream forking into separate read and write paths.

  • Architected a cloud-native IoT device orchestration platform for connected vehicles: real-time telemetry ingestion, multi-channel diagnostic event processing, routing to downstream analytics.
  • Delivered 20+ modular Spring Boot microservices; 80% system efficiency gain and per-device configuration management at scale.
  • Serverless incident-detection and alerting on AWS Lambda cut infra cost from ~$65,000 to ~$17,000 annually.
  • Designed 8 composable microservices (device registration, telemetry ingestion, rule evaluation, fleet analytics, alerting) — 65% infra cost reduction.
  • Angular + Spring Boot engineering dashboards for real-time telemetry, remote configuration, and field diagnostic workflows.
  • Project Reactor + CQRS to decouple read/write models — sub-second query response under peak telemetry load.
  • Ran FMEA across all platform components for failure-risk mitigation.
  • Java
  • Spring Boot
  • Angular
  • TypeScript
  • Kafka
  • AWS Lambda
  • DynamoDB
  • EC2
  • Docker
  • Oracle PL/SQL
  • Project Reactor
Ch. 04

Fluentgrid Limited

Technical Analyst, Utility Software Engineering

Oct 2011 – Jun 2015 · Visakhapatnam, India

A substation energising distribution feeders, an outage taking a branch offline, and a SCADA sweep detecting the fault before restoration propagates back down the line.

  • Built enterprise platforms for electricity distribution utilities: grid automation, DMS, outage management, SCADA-integrated tooling.
  • REST and SOAP APIs (Spring, ZK) integrating distribution modules, field device data, and back-office enterprise systems.
  • Data models for grid topology, meter data, outage events, and network segments supporting SCADA/DMS reporting across regional deployments.
  • Configuration and engineering workflow tools letting operations engineers define and automate grid-level processes.
  • Java
  • Spring
  • ZK
  • Oracle SQL
  • SOAP/REST
  • XML
  • DMS
  • SCADA

Software Engineer — TATA Consultancy Services · Jun 2015 – Nov 2016 · Security-focused team building high-security authentication for a financial client; TDD, SonarQube, vulnerability scanning, J2EE on WebLogic.

06 / Impact

Measured outcomes.

  • $65K → $17K

    Annual infrastructure cost, via serverless incident detection

    Robert Bosch

  • 80%

    System efficiency gain across 20+ modular services

    Robert Bosch

  • 65%

    Infrastructure cost reduction from 8 composable microservices

    Robert Bosch

  • 60%

    Reduction in inter-module dependencies after monolith decomposition

    GSK

  • 50%

    Improvement in transactional and reporting performance

    GSK

  • 40%

    Faster provisioning of new services via a scaffolding generator

    GSK

  • Sub-second

    Query response under peak telemetry load, via CQRS + Project Reactor

    Robert Bosch

  • Millions

    Daily transactions across the merchant lending platform

    U.S. Bank

Every figure above is drawn from the role it is attributed to. None of them were rounded up in their own favour.

07 / Projects

Selected case studies.

  • U.S. Bank

    Merchant Lending Platform

    Distributed microservice platform handling millions of daily transactions. Owned the architecture: service boundaries, data contracts, API governance via APIGEE, and a secure SDLC with threat modelling and CI/CD static analysis. Hybrid On-Prem/Azure on Rancher Kubernetes.

    Outcome

    Enterprise uptime SLAs met under chaos-tested failure conditions.

    • Java 17
    • Spring Boot
    • APIGEE
    • Kafka
    • Rancher/Kubernetes
    • Azure
  • GSK

    Enterprise Monolith Modernization

    Migrated Struts/WebLogic monoliths to Spring Boot microservices across a global platform, with centralized OAuth2 + Azure AD identity.

    Outcome

    60% fewer inter-module dependencies, 50% faster transactional and reporting performance, 40% faster new-service provisioning via a scaffolding generator.

    • Java
    • Spring Boot
    • OAuth2
    • Azure AD
    • Oracle PL/SQL
    • Jenkins
  • Robert Bosch

    Connected-Vehicle IoT Orchestration

    Cloud-native platform ingesting real-time telemetry from field-deployed devices, evaluating diagnostic rules, and routing to analytics. Serverless incident detection for safety-critical events. CQRS + Project Reactor for sub-second reads at peak load.

    Outcome

    Infra cost $65K → $17K annually; 20+ services; 80% efficiency gain.

    • Java
    • Spring Boot
    • AWS Lambda
    • DynamoDB
    • Kafka
    • Project Reactor
    • Angular
  • Fluentgrid

    Smart Grid Utility Tooling

    Distribution management, outage management, and SCADA-integrated engineering tools for electricity utilities. Grid topology and meter data modelling; REST/SOAP integration between field devices and back-office.

    Outcome

    Multi-zone regional utility deployments with real-time grid visibility for operations engineers.

    • Java
    • Spring
    • Oracle SQL
    • SOAP/REST
    • SCADA
    • DMS

08 / Capabilities

What I work with.

AI & Agents
  • Claude
  • GitHub Copilot
  • Agent Harnesses
  • Multi-Agent Orchestration
  • Skills & Personas
  • Rule Systems
Architecture
  • Microservices
  • Event-Driven
  • CQRS
  • DDD
  • Serverless
  • Edge-Cloud Hybrid
  • API-First
Backend
  • Java 8/11/17
  • Spring Boot
  • Spring Cloud
  • Spring Batch
  • Spring Security
  • Kafka
  • Python
Data
  • PostgreSQL
  • Oracle PL/SQL
  • MySQL
  • Hibernate/JPA
  • Redis
  • DynamoDB
  • Query Optimization
Cloud
  • AWS (Certified Solutions Architect)
  • Azure
  • GCP
DevOps
  • Docker
  • Kubernetes (Rancher/AKS)
  • Jenkins
  • GitHub Actions
  • GitLab CI
  • ArgoCD
  • Maven
Security
  • Secure SDLC
  • OAuth2
  • Threat Modelling
  • APIGEE Governance
  • SonarQube
  • DevSecOps
Frontend
  • Angular
  • React
  • TypeScript
  • HTML5/CSS3
Domain
  • SCADA/DMS/EMS Integration
  • IEC 61850 / CIM / IEC 62351
  • IoT Telemetry
  • Grid Automation

09 / Trajectory

Fifteen years, in order.

  1. Feb 2023 – Present

    Lead Software Development Engineer

    U.S. BankRole

  2. Dec 2021 – Feb 2023

    Software Architecture Lead

    GSKRole

  3. Nov 2016 – Nov 2021

    Software Development Engineer III, IoT & Connected Systems

    Robert BoschRole

  4. Jun 2015 – Nov 2016

    Software Engineer

    TATA Consultancy ServicesRole

  5. Oct 2011 – Jun 2015

    Technical Analyst, Utility Software Engineering

    Fluentgrid LimitedRole

  6. Dec 2011

    M.Tech, Computer Science

    JNTU KakinadaEducation

Education & Certification

  • M.Tech, Computer Science — JNTU Kakinada, Dec 2011 (7.2 CGPA, University Honor Society)
  • AWS Certified Solutions Architect