Imran Khan Full-stack engineer Resume

Full-stack engineer · 4+ years

Imran Khan

Full-stack engineer focused on observability, reliability, and useful software.

I own product work from React interfaces through FastAPI and Node.js services, async workers, AWS infrastructure, and the telemetry that keeps the whole system understandable in production.

Get in touch Resume GitHub LinkedIn

Experience

Four years across product, systems, and operations.

I work close to product and infrastructure boundaries: designing workflows, tracing failures, improving performance, and leaving systems easier to operate than I found them.

Aug 2023 - Present

Software Development Engineer 2

Procedure Technologies

Full-stack product and reliability ownership for a multi-tenant education platform used across the United States.

  • Led the production Datadog setup across React, FastAPI, Node.js, AWS, and background workers, including RUM, APM, logs, traces, dashboards, and alerts.
  • Instrumented three production services across API, worker, and EC2 workloads, reducing incident diagnosis from hours to approximately 30 minutes.
  • Used APM to isolate CSV reads responsible for about 20% of request latency, then reduced p95 latency by 75% through async processing and precomputation.
  • Designed APIs, data schemas, Lambda orchestration, async callbacks, and background processing with RabbitMQ and Dramatiq.
  • Managed AWS and MongoDB infrastructure with Terraform and participated in weekly on-call rotations backed by runbooks.
  • Built reusable workflows with Codex CLI, Claude Code, agent skills, MCP servers, terminal CLIs, and CI verification.
ReactTypeScriptFastAPINode.jsDatadogAWSTerraformRabbitMQDramatiqPandasPlaywright
Feb 2022 - Jul 2023

Software Engineer 1

MAQ Software

Frontend engineering for data-driven products with high daily usage and near real-time updates.

  • Built React and TypeScript dashboards serving thousands of daily active users.
  • Implemented Redux state flows and integrated APIs for near real-time updates.
  • Turned product requirements into maintainable frontend workflows through testing, review, and iterative delivery.
ReactTypeScriptReduxREST APIsHTMLCSS

Work in numbers

Context for the scale and outcomes behind my work.

A compact view of the users, production footprint, incident response, and performance improvements represented in my recent experience.

700+U.S. school districts supported by the multi-tenant platformProduct scale
5,000+Users across data-heavy planning and reporting workflowsUser reach
~30 minIncident diagnosis reduced from hours through correlated telemetryReliability
75% lowerp95 latency after isolating dependencies and moving work off-pathPerformance

How I work

The tools and habits I bring to a codebase.

A practical toolkit for startup environments where the same engineer needs to understand the user workflow, the distributed request, and the deployment path.

Observable systems

End-to-end telemetry and incident response across browser, API, workers, dependencies, and cloud infrastructure.

  • Datadog RUM, APM, logs, traces
  • RED metrics and dashboards
  • OpenTelemetry and Grafana LGTM
  • On-call and runbooks

Full-stack delivery

Product features that cross interface, API, data, queue, infrastructure, and automated test boundaries.

  • React and TypeScript
  • FastAPI, Node.js, Python
  • AWS and Terraform
  • RabbitMQ, Dramatiq, Pandas

AI-enabled engineering

Reusable, semi-automated workflows that accelerate analysis and implementation while retaining review and verification.

  • Codex CLI and Claude Code
  • Custom agent skills
  • MCP servers and terminal CLIs
  • Human review and CI gates

How I debug production

From alert to root cause, without guessing.

These are simplified reconstructions of two real investigations. They show the reasoning path and only use outcomes I can defend.

Debugging notes / anonymized production work Resolved
Investigation 01

Course request latency

A recurring slow endpoint narrowed from an alert to an inefficient database query.

  1. AlertFrequent course requests exceeded the expected duration.
  2. MetricsRED dashboards confirmed a recurring latency pattern.
  3. Trace + logsFollowed the request into its trace and correlated logs.
  4. Root causeThe database query loaded too much data without pagination.
  5. MitigationChanged the query path and validated the request end to end.
Investigation 02

CSV dependency bottleneck

APM made a hidden dependency cost visible enough to justify a system-level change.

  1. APM breakdownCSV reads accounted for approximately 20% of request latency.
  2. Design changeMoved heavy work off-path and precomputed reusable results.
  3. Outcomep95 latency decreased by 75% after the changes.

Writing

Notes on software engineering practice.

I write to make engineering ideas explicit: why a practice matters, where it helps, and how it changes day-to-day decisions.

Authored case study

The Architecture Shift That Took School Scheduling to 30+ Districts Seamlessly

How reusable SIS-level data pipelines reduced duplicate code, accelerated district onboarding, and enabled a zero-downtime migration.

 Testing practice

TDD: Write Quality Code, One Test at a Time

A practical look at Red-Green-Refactor, maintainable business logic, and the long-term value of test-driven development.

 Engineering craft

Evolution from Programmer to Software Engineer

How engineering judgment grows beyond writing code into documentation, goal-setting, and solution-oriented thinking.