Lakshay Walia Cloud, DevOps & AI Engineer
Cloud, DevOps & AI engineering

Scroll to connect the infrastructure and intelligence stack

Cloud infrastructure meets AI engineering. Built to ship, train, and operate.

I build secure cloud foundations, automate delivery, and develop applied AI workflows spanning model training, RAG, API serving, evaluation, and MLOps.

View connected work View résumé
  • Delhi, India
  • Remote collaboration
  • AWS · Kubernetes · AI/ML

Adaptive copy runs locally from the section you explore. No tracking.

Service routing online
EC2

Compute

Scalable application and workload hosting

VPC

Networking

Subnets, routing, security, and isolation

S3

Storage

Artifacts, backups, and data workflows

K8s

Orchestration

Containers, health, scaling, and GitOps

TRAIN

Model training

Data preparation, experiments, and evaluation

RAG

Retrieval systems

Embeddings, search, grounding, and APIs

MLOps

Model operations

Serving, versioning, monitoring, and rollout

CI/CD

Delivery automation

Build, test, publish, deploy, and verify

Deployment4h → 12m
Delivery modelRepeatable
Connected domainCloud + AI

Selected work

Evidence before adjectives.

Each project is labeled by context so client delivery, engineering labs, and open-source work are never presented as the same thing.

Client deliveryJuly 2026

Kobo Command Center

A customized data platform packaged as a repeatable 13-service deployment with generated secrets, health probes, system startup, backup workflows, and an operator handover.

Before
Approximately four hours of manual setup
After
Approximately twelve-minute automated deployment
  • Docker Compose
  • Django
  • PostGIS
  • Redis
Read case study Open simulation
Engineering lab2026

EduCenter OS

A cloud-native education platform used to demonstrate repeatable AWS provisioning, K3s orchestration, GitOps synchronization, and health-aware releases.

  • Argo CD synchronization and self-healing
  • Prometheus and Grafana monitoring path
  • Next.js and FastAPI application stack
  • AWS
  • K3s
  • Argo CD
  • FastAPI
Read case study Open simulation
Open engineering lab2026

LegacyOps 2026

A standardized delivery lab comparing Docker Compose, Kubernetes, and Helm workflows with automated container builds and readiness verification.

  • GitHub Actions and Jenkins delivery paths
  • GHCR container publishing workflow
  • Health, readiness, and HTTP verification
  • GitHub Actions
  • Jenkins
  • Helm
  • Docker
Read case study Open simulation

Looking for interactive experiments, architecture utilities, or incident simulations?

Visit the engineering lab

Expertise

Capabilities organized around outcomes.

Tools matter, but only when they improve delivery speed, resilience, security, or operational clarity.

01

Cloud foundations

AWS VPC design, IAM boundaries, secure networking, compute, storage, DNS, and recovery-aware infrastructure.

  • AWS
  • Terraform
  • Ansible
  • Linux
02

Deployment automation

Repeatable delivery paths that replace long checklists with versioned, testable automation.

  • Docker
  • Kubernetes
  • Helm
  • GitHub Actions
03

Reliability & operations

Health checks, observability, backup workflows, rollback planning, and documentation operators can use.

  • Prometheus
  • Grafana
  • Systemd
  • Runbooks
04

Applied AI systems

Hands-on AI engineering across dataset preparation, model training, retrieval systems, evaluation, API serving, and model operations.

  • Python
  • Model training
  • RAG
  • MLOps

Services

Clear scopes, practical handover.

Starting ranges are directional. A short discovery call confirms requirements, access, risk, and delivery timing.

Cloud foundation From ₹15K

AWS architecture & hardening

For teams that need a secure, understandable foundation before applications are deployed.

  • VPC and subnet architecture
  • IAM and security-group review
  • Backup and recovery baseline
  • Architecture notes and handover
Typical delivery: 1–2 weeks
Delivery automation From ₹18K

Container platform packaging

For multi-service applications that are slow, fragile, or inconsistent to install and operate.

  • Docker Compose or Kubernetes packaging
  • Generated configuration and secrets
  • Health checks and startup automation
  • Operator runbook and recovery steps
Typical delivery: 2–4 weeks
GitOps & reliability From ₹25K

Kubernetes delivery workflow

For teams ready to move from manual cluster changes to reviewed, observable delivery.

  • GitOps repository structure
  • Argo CD synchronization
  • Readiness and rollback strategy
  • Monitoring and handover session
Typical delivery: 3–5 weeks
Applied AI engineering Scoped after discovery

AI prototype to operated service

For teams moving an AI experiment beyond a notebook into a testable, observable application workflow.

  • Dataset preparation and model-training workflow
  • RAG pipeline, embeddings, and retrieval API
  • FastAPI model serving and container packaging
  • Evaluation, versioning, monitoring, and handover
Typical delivery: confirmed after technical discovery
Every engagement includes Written scopeVersion-controlled deliveryVerification checklistHandover documentation

Working process

Designed to reduce surprises.

Access, success criteria, and rollback expectations are agreed before implementation begins.

  1. 01

    Discover

    Map the current system, constraints, owners, and failure points.

  2. 02

    Design

    Agree on architecture, risks, deliverables, and acceptance checks.

  3. 03

    Build & verify

    Implement in version control and validate the operational path.

  4. 04

    Handover

    Transfer documentation, recovery steps, and ownership clearly.

LW Delhi · India

About

I like systems that explain themselves.

I’m Lakshay, a cloud, DevOps, and AI engineer based in Delhi. My work connects infrastructure, delivery automation, model workflows, and the people who operate what gets delivered.

I care about predictable deployments, useful health signals, realistic recovery plans, and documentation that survives the original builder. My AI engineering work includes model-training experiments, retrieval systems, evaluation, API serving, and MLOps workflows.

CurrentIT internship · Government IT divisionJune 2026–present
ExperienceFreelance business software supportAugust 2024–May 2026
EducationBCA, GGSIPU · MCA, IGNOUMCA in progress, expected 2028
Open résumé LinkedIn GitHub Instagram

Technical writing

Decisions, not just finished screens.

Notes on deployment design, GitOps workflows, and infrastructure trade-offs.

Deployment automation · 8 min

Packaging a 13-service stack into one repeatable install

Dependency ordering, generated secrets, health checks, and the operator experience.

Read article GitOps · 7 min

What self-healing actually means in an Argo CD workflow

Repository structure, drift correction, readiness, and when automation should stop.

Read article Infrastructure as code · 6 min

Designing a reusable AWS VPC module without hiding risk

Inputs, boundaries, safe defaults, cost decisions, and output verification.

Read article

Contact

Bring the messy deployment.

Share the current state, what keeps breaking, and what “done” should look like. I’ll reply with questions or a realistic next step.

Response targetWithin two business days
TimezoneIndia Standard Time · UTC+5:30
Add a number only if you would like Lakshay to call you.

Your inquiry is sent to Lakshay and stored securely for follow-up. Your phone number is optional and is never displayed publicly.