Aniket SinghResume

Currently

Junior Developer at Florosense ESPL · Bengaluru, Karnataka

Junior Developer at Florosense — shipping IoT dashboards, field-service platforms, and regulatory data pipelines.

I work as a Junior Developer at Florosense ESPL in Bengaluru, building production software for ambient air-quality monitoring. That includes Spring Boot microservices, a Node.js and Next.js client portal, scheduled integrations with DPCC, and the Docker/Nginx deploys that keep those systems online. Earlier I interned at Probuddy Software Solutions on Eureka, API Gateway, and REST APIs. I also maintain a consistent DSA practice (350+ problems).

View selected workGet in touch

Microservices · IoT telemetry · RBAC dashboards · Docker / Nginx

Aniket Singh, Junior Developer at Florosense ESPL

Career

Experience

  1. Feb 2026 — Present

    Bengaluru, Karnataka

    Current role

    Junior Developer · Florosense ESPL

    Full-time role across Florosense’s air-quality product line: field operations, regulatory posting, and client-facing analytics.

    • Own Digital EForm end to end — React SPA, five Spring Boot microservices, JWT RBAC, PostgreSQL, and HTTPS delivery through Docker Compose and Nginx at digitalform.florosense.com.
    • Built Sensor Bridge, a scheduled Spring Boot integration that reports live PM2.5/PM10 to the DPCC portal every 15 minutes with idempotent posting (cooldown + lock).
    • Contribute to Duton Software: Node.js APIs and a Next.js dashboard that store sensor telemetry, client records, and site locations, with role-based feature access for each client account.
    • Design REST contracts, IoT data mappings, and VPS releases for the production air-quality stack.

  2. May 2025 — Jan 2026

    Bengaluru, Karnataka

    Software Developer Intern · Probuddy Software Solutions

    Backend internship focused on Spring Cloud microservices and the SDLC around them.

    • Implemented REST APIs on Spring Boot and MySQL using layered architecture (controller, service, repository) and Bean Validation DTOs.
    • Configured Eureka service discovery and Spring Cloud Gateway for authenticated, path-based routing across services.
    • Participated in Agile sprints: code reviews, deployments, and release checklists.

Selected work

Projects from the Florosense stack

Four systems I help design, build, and run: field e-forms, the client sensor portal, DPCC posting, and a 7-day air-quality forecast model.

01Live · Florosense

Digital EForm

Digital Installation & PM Visit E-Form System

Field engineers at Florosense used paper checklists for pre-visit, installation, calibration, and preventive maintenance. Digital EForm replaces that workflow with a React single-page application backed by five Spring Boot microservices (auth, pre-visit, installation, calibration, PM). Each service owns its schema in a shared PostgreSQL database. Authentication issues a JWT (JJWT, BCrypt). Sister services share the same JwtAuthFilter and signing secret, so one session covers the platform. USER accounts create and view reports; ADMIN accounts can edit and delete.

How it is built

  • Preventive maintenance is a six-step wizard. Client state is submitted once as JSON to POST /api/pm_reports. Checklist lines persist as child entities across PHYSICAL_INSPECTION, POWER_SUPPLY, SENSOR_HEALTH, COMMUNICATION, CALIBRATION_PERFORMANCE_VERIFICATION, and CLEANING_ACTIVITY.
  • Pre-visit and installation accept multipart site photos after the report id exists. Images are stored as PostgreSQL bytea and returned as Base64 — no public object store.
  • Report numbers are generated in-service (PM-YYYY-XXXX, FESPL_CAL_YYYYMMDD_XXXX). Calibration records before/after readings against a master instrument, with due dates defaulting to +90 days and UUID primary keys.
  • The dashboard aggregates /count from each service (two-hour cache) and polls notifications every 20 seconds. Completed reports export to A4 PDF through html2canvas and jsPDF.
  • Production runs on Docker Compose. Nginx is the public edge: SPA on /, APIs under /api/*, TLS at digitalform.florosense.com. A Spring Cloud Gateway module exists for local routing; production traffic does not use it.
  • React 19
  • Spring Boot 3.5
  • Spring Security
  • JWT / RBAC
  • PostgreSQL
  • JPA / Hibernate
  • Docker Compose
  • Nginx TLS
Open live productView source
02Live · Florosense

Duton Software

Client sensor platform · location, telemetry, and RBAC dashboards

Duton is Florosense’s client-facing operations product. We ingest and manage full sensor records — device telemetry, associated client accounts, and site location — then expose that data through a Next.js dashboard. Each client logs in to their own view. Role-based access control limits which modules, sensors, and actions they can use, so an admin, operator, and read-only stakeholder do not see the same surface.

How it is built

  • The API layer is Node.js. It handles authentication, client and site CRUD, sensor registry, and time-series reads for PM and environmental parameters.
  • The UI is Next.js: authenticated routes, dashboard widgets for live and historical readings, and maps/lists bound to stored location data for each installation.
  • Client onboarding stores organisation profile, contact, and geo coordinates for every site so operators can find a node by place, not only by device id.
  • Feature flags follow the user’s role. Higher-privilege roles get configuration, user management, and export; restricted roles get monitoring views only.
  • Duton sits on the same production VPS as Sensor Bridge and Digital EForm, so ingest, reporting, and field ops share one operational footprint.
  • Node.js
  • Next.js
  • REST APIs
  • RBAC
  • Geolocation
  • IoT telemetry
  • PostgreSQL
03Live · Florosense

Sensor Bridge

DPCC regulatory integration · scheduled PM2.5 / PM10 posting

The Delhi Pollution Control Committee (DPCC) requires PM2.5 and PM10 from our sites on a 15-minute cadence. Sensor Bridge is a Spring Boot worker that pulls the latest reading from the Florosense cloud, resolves sensor id to DPCC site id, and POSTs the payload. In production, a scheduler tick overlapping a retry could double-submit. A cooldown window and a lock around the outbound call make posting idempotent.

How it is built

  • A fixed-rate scheduler fires every 15 minutes. Before HTTP, the worker checks last-success time and an in-process lock; if either says wait, the tick exits.
  • Sensor-to-site mappings live in the database so operations can register a new node without a rebuild.
  • Operational APIs cover config, health, recent logs, and a manual trigger for a single sensor when DPCC requests a backfill.
  • Shipped as the sensor-bridge-api image with Compose restart policies on the Hostinger VPS alongside Duton and EForm.
  • Spring Boot
  • @Scheduled
  • REST client
  • Idempotent jobs
  • PostgreSQL mappings
  • Docker
  • Actuator
04ML · Florosense

Future Sensor Prediction

LSTM and XGBoost · 7-day site workability and PM forecast

Construction sites needed a forward look: can work continue over the next seven days, and how will PM2.5/PM10 move? I treated that as binary classification plus multivariate regression. Features come from a timestamp-aligned merge of on-site sensors (PM, temperature, humidity) and CPCB Mulund West meteorology (NO₂, O₃, CO, RH, wind, solar radiation, rainfall, ambient temperature).

How it is built

  • Anomaly cleaning first (spikes, dead sensors), then lag engineering: PM and weather at t-1, t-2, and t-3.
  • Compared LSTM sequence models (.h5) with XGBoost regressors/classifiers (.pkl) on the same splits rather than isolated notebooks.
  • Residuals between sensor and model were explained with feature importance — wind and nearby construction activity were frequent drivers.
  • Inference is served through FastAPI in front of the pickled models and scalers so a dashboard can request a 7-day workability flag without opening Jupyter.
  • Python
  • LSTM / Keras
  • XGBoost
  • Time-series features
  • FastAPI
  • Jupyter
View source

Academics

Education

Aug 2022 — July 2024

Master of Computer Applications

Makhanlal Chaturvedi National University of Journalism and Communication

Bhopal · CGPA 7.8

Certification

Java Full-Stack Developer

JSpiders, Bengaluru

Certification

Java Developer Associate

ISO 14001

Capabilities

Skills

Languages

  • Java 17
  • JavaScript
  • Node.js

Backend

  • Spring Boot 3
  • Spring Security / JWT
  • Spring Cloud (Eureka, Gateway, OpenFeign)
  • Hibernate / Spring Data JPA
  • Express / Node.js APIs
  • Apache Kafka

Frontend

  • Next.js
  • React
  • Vite
  • Axios

Data

  • PostgreSQL
  • MySQL
  • Redis
  • NoSQL

Ops

  • Docker Compose
  • Nginx
  • Jenkins
  • Maven
  • Git

Let’s talk

Open to Java, Node.js, and full-stack conversations.

Based in Bengaluru. Email is the fastest path if you want to discuss a role or a collaboration.

Emailankitparihartu98@gmail.comPhone+91 6264536235GitHubgithub.com/aniketSingh4LinkedInaniket-singh-831388273