I'm Devathmaj.
I build for
the long run.
Software engineer, working from database to browser.
I'm a computer science graduate (2026) based in Kerala, India. I work across the stack: designing schemas and REST APIs, building the interfaces that sit on top of them and shipping the result in containers on Azure.
The work is technical.
The intention is simple:
make it dependable.
Most of my work has been about making operational systems trustworthy. At Kannur Vision I built a full-stack distribution management system with audit logging and role-based access across eight hierarchical roles. At Kerala Vision Broadband I built a network backup platform that automated device backups, configuration management and log comparison, cutting manual network administration effort by 80%.
I care about observability, reliability, platform and delivery: systems that can be measured, operated and kept healthy under real load. My projects have included hybrid semantic search with pgvector, encrypted real-time chat over WebSockets and Redis, AI extraction pipelines with Groq and Gemini and Azure deployments with Key Vault and Content Safety.
My background is a B.Tech in Computer Science and Engineering (2022–2026) and I'm looking for a team where I can work on backend, infrastructure, reliability or platform problems.
- Focus
- Backend Engineering · Cloud & DevOps · Automation · AI Integration
- Education
- B.Tech CSE, 2022–2026
- Based in
- Kerala, India
- Status
- Open to software, backend and infrastructure roles
Good systems.
Every layer considered.
From the data underneath to the experience on top. Explore the parts of the stack I work with.
Start with a solid foundation.
Schemas, access control and search. My work includes PostgreSQL data models, audit logging and hybrid semantic search with pgvector.
PostgreSQL · SQLAlchemy · pgvector · Redis Explore ConfessItMake the complexity usable.
REST APIs and the interfaces that sit on top of them. I connect React frontends to Python services, including authenticated and real-time experiences.
React · TypeScript · FastAPI · WebSockets Explore the distribution systemBuild it. Then ship it.
Containers, cloud deployments and automation. My projects use Docker and Azure, with Key Vault and automated delivery in the stack.
Docker · Azure · GitHub Actions · Linux Explore VoucherBotKeep the system understandable.
Systems should be measurable and operable. My focus includes observability, backup automation, configuration management and log comparison.
OpenTelemetry · Prometheus · Grafana · Loki Explore the backup platformSelect a layer to look a little closer.
Experience
The work behind
the way I work.
Selected projects
Two projects that show the kind of systems I like to build.
ConfessIt
Anonymous social, gaming & matchmaking platform
A full-stack anonymous platform with magic-link and JWT auth, rotating refresh tokens, hybrid semantic search, encrypted real-time chat and nine social games.
- Python
- FastAPI
- PostgreSQL
- pgvector
- Redis
- React
- Docker
- Azure
VoucherBot
Automated deal discovery & notification system
An async service that monitors 30+ sources, extracts deal details with an AI pipeline and Gemini fallback and delivers idempotent Discord and Telegram notifications.
- Python
- FastAPI
- PostgreSQL
- SQLAlchemy
- Docker
- Supabase
- Cloudflare Pages
- Groq
- Gemini
Toolkit
The tools behind the work and the part each one plays. Open a group, then choose a tool to look closer.
APIs & interfaces Services, contracts & screens
From a validated request to an interface someone can actually use.
FastAPI: Give the interface a reliable API.
Python endpoints handle requests, authentication and business rules. FastAPI is a core part of my backend stack for full-stack applications.
Pydantic: Make the data contract explicit.
Typed models validate incoming payloads and describe expected data. They keep validation close to the API instead of scattering checks through the application.
React: Turn services into usable screens.
Components connect application state, user input and API responses. I use React with TypeScript to build the interfaces above my Python services.
Also in the mix
In practice: distribution management
- Python
- Flask
- TypeScript
- JavaScript
- JWT
- RBAC
Data & real-time systems Store, discover & communicate
Different data problems need different tools, even within the same product.
PostgreSQL: Keep a durable source of truth.
Relational schemas hold structured application data. In ConfessIt, PostgreSQL keeps the durable chat transcript while live messages travel through Redis.
pgvector: Search beyond an exact keyword.
Vector similarity adds a semantic dimension to search inside PostgreSQL. ConfessIt uses pgvector as part of its hybrid search approach.
Redis: Move messages while they matter.
Redis pub/sub distributes live chat messages between connections. It complements the persistent database rather than replacing the conversation record.
Also in the mix
In practice: ConfessIt
- SQL
- SQLAlchemy
- WebSockets
- MySQL
- MongoDB
- Supabase
Cloud & delivery Package, automate & deploy
The tools that take a service from a local build to a cloud deployment.
Docker: Ship the environment with the service.
Container images package an application and its dependencies together. I use Docker to prepare services for deployment rather than relying on a machine-specific setup.
GitHub Actions: Make delivery repeatable.
Repository workflows provide a place to automate checks, builds and delivery steps. GitHub Actions is part of my cloud and delivery toolkit.
Azure: Give the application a production home.
My projects deploy containerized services on Azure. Supporting services include Key Vault for secrets and Content Safety for content moderation.
Also in the mix
In practice: deploying ConfessIt
- Linux
- Nginx
- AWS
- Cloudflare Pages
- Key Vault
Automation & observability Verify, instrument & understand
Testing and observability make system behavior easier to reason about.
Pytest: Make expectations executable.
Python tests turn expected behavior into repeatable checks. Pytest is my testing tool for exercising service logic and catching regressions.
OpenTelemetry: Follow work through a system.
Instrumentation adds traces and telemetry around application behavior. OpenTelemetry is part of my focus on systems that can be measured and operated.
Grafana: Make the signals readable.
Dashboards bring operational signals into a view that can be investigated. Grafana sits alongside Prometheus and Loki in my observability toolkit.
Also in the mix
In practice: network backup automation
- Selenium
- Prometheus
- Loki
- Netmiko
- Rclone
- SSH
- Groq
- Gemini
- Azure AI
- Azure Content Safety
Education & credentials
The degree behind the work and the credentials earned alongside it.
St. Thomas College of Engineering & Technology, Mattanur
B.Tech in Computer Science
Coursework Data Structures & Algorithms, Operating Systems, Database Management Systems, Computer Networks, Management of Software Systems, OOP.
Selected certifications
Let's build something
dependable.
Open to software, backend and infrastructure roles. Have something in mind? I'd like to hear about it.
Visit the contact page