Skip to content
Devesh Singh
← All work

AIOps

NextDecade Observability

Project LeadJun 2026 – Jul 2026

Starting point

Started from an empty repository

Nothing was here before — I set up the repository and built the whole thing.

What I owned

  • Designed
  • Led
  • Built
  • Shipped
  • Operated

The moving parts

What talks to what — clients, services, data, and the boundaries between them.

IdentitySign-in providers
Monitoring UIClients
IngestionExternal source APIs
BackendServices
AI analysisThird-party APIs
Raw log payloadsData
DatabaseData

Push to prod

The checks a change has to pass before it ships, and what watches it after.

Source
CI/CD
Build
DeployFE · BE · Multi-env

Where it lives

The cloud it runs on — ingress, compute, data and messaging — and how that gets provisioned.

Azure infrastructure
Infra as codeDelivery
ComputeInfrastructure
Object storageData
Managed dataData
AI modelsManaged inference

Node detail

Pick a box

Tap any box in the diagrams to see the tools behind it and what each one did on this project. Tap again to let go.

The product

AI-assisted log analysis and monitoring platform for a sustainable-energy company, part of a wider operations-automation programme.

What I built

Project lead on a fixed-scope, 6-week engagement with a team of 4 developers and 1 QA — mostly backend and DevOps work. Owned system design, code review and delivery; built the log ingestion and analysis platform on Azure App Service, Cosmos DB and Blob Storage, wired GPT-5.4 through Azure AI Foundry to explain failures, wired sign-in through Okta federated with Microsoft Entra ID, governed privileged access with Microsoft Entra PIM, and provisioned the infrastructure with Terraform from Azure Repos.

Everything I did here

  • Project lead on a fixed-scope engagement (6 weeks) with a five-person team — 4 developers (2 frontend, 2 backend) and 1 QA — owning system design, code review and delivery, working mostly on backend and DevOps.
  • Built an AI-assisted log analysis and monitoring platform that surfaces issues and suggested fixes directly from application logs.
  • Integrated logs and metrics from multiple sources through their APIs into one view.
  • Used GPT-5.4 through Azure AI Foundry to analyse logs across the stages of the process lifecycle, producing plain explanations and suggested fixes so teams resolve incidents faster.
  • Provisioned Azure infrastructure with Terraform — App Service for the platform, Cosmos DB for stored logs and analyses, Blob Storage containers for raw log payloads.
  • Integrated single sign-on: Okta as the identity provider, federated with Microsoft Entra ID, so engineers reach the platform with their existing corporate account and no separate password.
  • Governed privileged access to the Azure subscription with Microsoft Entra Privileged Identity Management, so elevated roles are requested and time-bound rather than standing.
  • Source control and pipelines run out of Azure Repos and Azure DevOps.
  • Set up CI/CD pipelines for the MERN stack.
  • Delivered as part of a larger programme to fully automate operations for a sustainable-energy solutions company.
Download PDF