d.Dinesh Kumar GENGINEER · BUILDER · MENTOR
DINESH KUMAR G BENGALURU, INDIA · BUILDING GLOBALLY
Staff Software Engineer & AI Lead

Engineering ideas.
Into real impact.

I turn complex problems into thoughtful products.
Building agentic AI, cloud platforms, and the teams behind them.

CURRENTLY AT DILIGENT PREVIOUSLY AT MANUFACTONTHOUGHTFUL SYSTEMS. MEASURABLE OUTCOMES. ↓
~93%↓

AI screening cost

1,500+

automated test cases generated

25K+

organizations provisioned

5×

consecutive Hackfest wins

01 / SELECTED WORK

From idea to impact.

A few things I’ve helped bring
into the world at Diligent.

AGENTIC AI · PRODUCT ENGINEERING

Intelligence, from zero to one.

Taking an AI-native risk intelligence platform from concept to enterprise production. Multi-agent retrieval and reasoning, designed around real customer workflows.

Amazon BedrockRAGTypeScriptAWS CDK
~93%reduction in AI screening cost
Behind the work +

Led architecture and delivery through beta, general availability, and enterprise rollout. Redesigned agent workflows and used deterministic search where appropriate to reduce screening costs from $8 to $0.55. Built an evaluation framework using approximately 5,000 human-annotated records.

DEVELOPER EXPERIENCE · MULTI-AGENT SYSTEMS

Less repetition. More engineering.

A multi-agent workflow that turns structured manual test cases into Playwright tests, with autonomous repair and deterministic validation.

Playwright MCPStrands AgentsZodTypeScript
1,500+end-to-end test cases generated
Behind the work +

Built the initial framework in three days. Specialized generation, repair, and evaluation agents work alongside TypeScript compilation, schema validation, and Playwright verification to create reliable automation.

CLOUD ARCHITECTURE · PLATFORM ENGINEERING

Built for the next chapter.

Leading a business-critical platform from Azure to AWS, spanning infrastructure, application architecture, databases, and delivery pipelines.

AWSEvent-driven systemsIaCCI/CD
$60Kannual infrastructure savings
Behind the work +

Led the end-to-end migration of the Third-Party Management product. Replaced legacy cron infrastructure with resilient serverless scheduling, established AWS-native patterns, and mentored engineers through architecture reviews and cross-team delivery.

02 / MORE FROM THE WORKBENCH

Built to make building easier.

Platforms and tools that
multiply a team’s impact.

PLATFORM ENGINEERING

A better path to provisioning.

Architected Alt-SDK, an internal alternative for enterprise GRC provisioning, with a real-time WebSocket CLI and a custom Go-based Terraform provider.

2×performance gains through execution ordering and parallelism
GoTerraformWebSocketsAWS CDK
Inside the project +

Led eight engineers delivering changeset generation, dependency-aware execution, and child-resource parallelism. A Fargate proof of concept demonstrated 3× performance improvement; a reusable CDK WebSocket construct was adopted by three engineering departments.

EVENT-DRIVEN SYSTEMS

Onboarding, without the bottleneck.

An event-driven pipeline using EventBridge, SQS, and Lambda to provision enterprise organizations and users with zero manual intervention.

1M+users provisioned in under three hours, alongside 25K+ organizations
EventBridgeSQSLambda
Inside the project +

Designed the pipeline to scale horizontally and support large enterprise rollouts. Automated provisioning brought customer onboarding into the platform’s application lifecycle rather than relying on manual operations.

03 / THE PERSON BEHIND THE WORK

Deep in the details.
Eyes on the bigger picture.

I’m a Staff Software Engineer and AI Lead at Diligent, working at the intersection of agentic AI, distributed systems, and product engineering.

Over the last eight years, I’ve taken products from a blank page to enterprise rollout, modernized cloud platforms, and helped engineering teams move with clarity and confidence.

I care about pragmatic architecture, thoughtful developer experiences, and making complicated things feel simple.

SEP 2021 — PRESENT

Diligent ↗

Staff Software Engineer · AI Lead Engineer

Architecting AI-native products and cloud platforms, from 0→1 delivery through enterprise rollout.

  • Technical leadership and mentorship across a 15-member engineering team.
  • Product, sales, and executive partnership to align architecture with customer needs.
  • AI evaluations, security readiness, and infrastructure ownership across the product lifecycle.
SEP 2018 — SEP 2021

Manufacton

Software Engineer

Built enterprise web and mobile applications for construction manufacturing and supply-chain workflows.

  • Led a four-member team owning the back-office ecosystem.
  • Delivered a one-month AngularJS to Vue.js migration.
  • Built APIs, analytics integrations, automated backups, and a logistics scheduling engine.
Get the full résumé ↓
04 / THE TOOLKIT

Depth where it matters.
Range where it helps.

Architecture to implementation.
Infrastructure to interface.

01

Applied AI

From retrieval to reasoning, with cost, quality, and evaluation built into the workflow.

  • Amazon Bedrock
  • Strands Agents
  • RAG & knowledge bases
  • MCP integrations
  • AI evaluations
  • LangGraph
02

Systems & cloud

Resilient platforms, clear domain boundaries, and infrastructure that teams can own.

  • Event-driven architecture
  • AWS CDK & Terraform
  • Lambda & Fargate
  • PostgreSQL & Redis
  • Microservices
  • CI/CD
03

Product engineering

Full-stack delivery and developer tools that move an idea all the way into production.

  • TypeScript & Go
  • React & Next.js
  • Node.js & NestJS
  • Playwright
  • Platform engineering
  • Technical mentorship
05 / RECOGNITION & FOUNDATIONS

A few milestones along the way.

THE FOUNDATIONB.E. in Computer ScienceNagarjuna College of Engineering & Technology2014 — 2018
THE FULL PICTURE, IN ONE PLACE

Prefer the résumé?

Experience, projects, skills, and the details behind the work.

Download résumé ↓
06 / FIELD NOTES

Ideas worth unpacking.

All writing
Agentic AI / Architecture

AI cost is an architecture problem.

Why the most effective optimization sometimes means making fewer model calls, rather than finding a cheaper model.

Developer Experience / Agentic AI

Reliable automation needs more than generation.

A practical way to think about AI-generated tests: generation, repair, and evaluation as separate responsibilities.

HAVE SOMETHING IN MIND?

Let’s build
something meaningful.

dineshwork2018@gmail.com