Backend developer
APIs, integrations and background jobs that keep working when the systems around them don’t.
- PHP
- Python
- REST APIs
- Webhooks
- Queues
Open to backend and server engineering roles
Backend developer Server engineer WordPress engineer Performance engineer Automation builder
I build the backend, run the Linux servers it lives on, and write the tools that keep both healthy. At BrightRays I look after 100+ production websites and apps, end to end.
0requests served
0%answered from cache
0bad requests blocked
0production websites and apps I look after
0databases tuned, audited and backed up nightly
0user-facing downtime per release
0uptime target across the fleet
One engineer, many hats
In a lean team you don’t get to stay in one lane. These are the roles I cover day to day.
Server and infrastructure engineer
Linux servers, web servers and PHP runtimes, sized and tuned so they handle traffic without needing bigger hardware.
- Linux
- Nginx
- OpenLiteSpeed
- PHP runtimes
- Capacity
WordPress engineer
Custom plugins, REST endpoints and deep fixes for WordPress running at fleet scale, from page builder to SQL.
- Plugins
- Hooks
- REST API
- WP-CLI
- Elementor
Database administrator
Keeping MySQL and MariaDB fast and lean: memory sizing, query and index work, growth control and safe backups.
- MySQL
- MariaDB
- Redis
- Backups
Performance engineer
Core Web Vitals, server response times and caching layers, measured before and after so the gains are real.
- Core Web Vitals
- Caching
- Lighthouse
- CDN
Security and reliability
Hardening servers and apps, filtering abusive traffic, and owning incidents through to a written root cause.
- Firewalls
- SSH
- Headers
- Postmortems
Tooling and automation
Monitors, audit tools, CI/CD pipelines and AI agent integrations that replace manual checks and paid tools.
- Cloudflare Workers
- GitHub Actions
- Cron
- MCP
Technical SEO
Making sites easy to crawl and index: redirects handled at the server, canonicals, sitemaps and structured data.
- Crawlability
- Redirects
- Schema
- Sitemaps
How I run servers
Every server I manage goes through the same loop. It’s how 100+ production apps stay fast, safe and boring, in the good way.
1
Set up and harden
A new server starts locked down, not patched later.
- Key-only SSH, firewalls and brute-force protection from day one
- Web server, PHP runtime and database installed to a known baseline
- Security headers and blocks on sensitive files at the server level
2
Tune the stack
Default settings are built for one small site. Shared servers need more care.
- PHP worker pools and opcode cache sized to the real workload
- Database memory sized to keep hot data in RAM
- Page, object and edge caching layered so most requests never reach PHP
3
Watch everything
I find out about problems before clients do.
- Uptime, SSL and response checks across the fleet
- CPU, memory, swap and disk alerts with sensible thresholds
- Speed scores tracked over time, so slowdowns show up as a trend
4
Respond and find the real cause
Restarting a service buys minutes. Finding the cause stops it coming back.
- Live triage of processes, logs and slow queries
- Fixes made backup-first, checked before they touch live data
- A short written root cause so the whole team learns from it
5
Ship without downtime
Releases should be boring for users.
- Atomic releases that switch over in one step
- Automatic health checks after every deploy
- Instant rollback if anything looks wrong
6
Back up and recover
A backup only counts if it restores.
- Nightly automated database and file backups, without locking tables
- Archives verified and copied offsite with rotation
- Recovery steps written down and tested
Problems I solve
The kinds of trouble I’m usually called in for, and how I handle each one.
Slow sites and overloaded servers
What it looks like
Pages take seconds to start loading, admin screens crawl, and server load climbs at peak hours.
How I handle it
Find where the time goes, then fix it at the right layer: worker pools, opcode and database memory, caching, and front-end weight.
What you get
Fast responses and calm servers on the same hardware, with no upgrade bill.
Server loadBefore
Crashes, memory exhaustion and server errors
What it looks like
White screens, failed saves, gateway timeouts and errors that only happen under real traffic.
How I handle it
Trace the failure through logs, memory use and the data itself until the real cause shows, instead of just raising limits.
What you get
The error stops for good, and a check goes in place so the same pattern gets caught elsewhere.
Error rateBefore
Databases that keep growing and slowing down
What it looks like
Queries get slower each month, disks fill up and backups take longer and longer.
How I handle it
Audit every database for bloat and missing indexes, size memory properly, and clean up safely after a backup.
What you get
Lean, fast databases, with limits in place so they stay that way.
Database sizeBefore
Risky, manual deployments
What it looks like
Code copied onto live servers by hand, broken pages mid-deploy, and bugs found by clients first.
How I handle it
Build each release separately, switch over in one atomic step, run health checks, and roll back automatically on failure.
What you get
Releases with no user-facing downtime, and a safe way back if anything goes wrong.
Availability during releasesBefore
Security gaps and abusive bot traffic
What it looks like
Exposed files, brute-force attempts, scanners hammering the site, and missing security headers.
How I handle it
Harden access, block exploit paths and junk traffic at the web server, and add security headers across every site.
What you get
Smaller attack surface, and bad traffic stopped before it costs a single PHP request.
Junk traffic reaching the appBefore
Repetitive manual operations
What it looks like
People checking sites by hand, copying backups, and paying for tools that only do part of the job.
How I handle it
Turn the routine into scheduled jobs, monitors and small internal apps the team can rely on.
What you get
Hours back every week, fewer missed issues and lower subscription costs.
Hours spent on routine checksBefore
Sites search engines struggle with
What it looks like
Wasted crawl budget, redirect chains, duplicate pages and poor Core Web Vitals.
How I handle it
Handle redirects and dead URLs at the server, fix canonicals and sitemaps, add structured data and improve speed.
What you get
Sites that are faster to crawl, cleaner to index and better for real visitors.
Pages crawled and indexed cleanlyBefore
Tools I’ve built for the team
When a job comes up every week, I build something so nobody has to do it by hand again.
Performance monitoring
SpeedPulse
Tracks Lighthouse scores and real-user Core Web Vitals for groups of websites on a schedule, and keeps the full history so slowdowns show up as trends.
- Runs entirely on Cloudflare’s free tier
- Scheduler and queues spread work inside API and CPU limits
- Charts drawn in plain SVG, no chart library
TypeScriptCloudflare WorkersD1QueuesKV
Fleet monitoring
WatchTower
A serverless monitoring system that checks uptime, SSL expiry, response codes and server health across the whole fleet, and alerts the team within a minute of trouble.
- Separate workers for checks, ingestion, processing and alerts
- Spots unusual trends, not just hard failures
- Replaced a paid third-party uptime service
Cloudflare WorkersQueuesD1KVWebhooks
Database operations
Database audit tool
Scans every database on a server for bloat, oversized tables, orphaned data and fragmentation, then cleans up safely in batches without downtime.
- One run covers every database on a shared server
- Clear report of what’s wasting space and why
- Safe cleanup with backups taken first
BashSQLMariaDBMySQL
Content quality
Blog QA Checker
A web app the editorial team uses to check posts before they go live. It gives each post a score and a clear list of fixes.
- Ten rule-based checks run first, in plain Python
- AI is only asked about what the rules flagged, and its answers are validated
- Catches broken links, heading problems, missing alt text and metadata
PythonFastAPISQLiteLLM APIs
AI agent integrations
Custom MCP servers
Servers that let AI coding assistants work with real engineering data safely, so they can run performance audits and coordinate changes across sites.
- Performance MCP runs Lighthouse audits and response-time checks from many locations
- Site memory MCP keeps a shared change log with locks, so agents don’t overwrite each other
- Works with Claude, Cursor and Antigravity
Model Context ProtocolTypeScriptPythonlibSQL
What I work with
I pick the tool that fits the problem. These are the ones I reach for most.
PHP 8PythonTypeScriptJavaScriptSQLBashWordPressSymfonyFastAPIReactNext.jsREST APIsWebhooks
LinuxNginxOpenLiteSpeedApacheMariaDBMySQLRedisDockerGitHub ActionsCloudflareAWSDigitalOceanLighthouse
Build
Backend services, APIs, WordPress plugins and internal web apps in PHP, Python and TypeScript.
Run
Linux servers, web servers, PHP runtimes, databases, caching, backups and deployments.
Improve
Speed, security, crawlability and automation, measured before and after.
Experience
January 2026 to now
PHP Backend Developer
BrightRays, Noida
- Own backend and infrastructure for 100+ production websites and apps on shared and dedicated Linux servers.
- Build plugins, APIs and integrations, and look after 70+ MySQL and MariaDB databases.
- Handle performance, security, technical SEO, deployments and incident response.
- Built the team’s internal tools for monitoring, auditing and content quality.
October 2024 to December 2025
Freelance and Contract Web Developer
Remote
- Built and launched websites, dashboards and small APIs for business clients.
- Owned each project from requirements to launch, working directly with clients.
- Worked across PHP, WordPress, React and TypeScript.