Available for ambitious systems workIndia · working globally

IoT × backend × edge intelligence

From field signals to intelligent systems.

I'm Anand Kumar Maurya, an IoT and backend engineer who turns messy real-world signals into reliable software, useful interfaces, and AI-assisted decisions.

For 4+ years I have shipped production systems across industrial automation, device tracking, Spring Boot microservices, Linux infrastructure, edge AI, and agentic workflows.

Download Resume

View Featured Projects Explore My Homelab

Selected toolchain

JavaSpring BootMicroservicesSystem DesignClean ArchitectureSpring SecurityOAuth2/OIDCKeycloakJWTRabbitMQRedisPostgreSQL

{"topic":"gate/entry","rfid":"E200341201","status":"granted"} // {"topic":"yard/container","id":"MSCU-2201847","gps":"22.312,70.802"} // {"topic":"weighbridge/tx","gross_kg":42180,"tare_kg":14200,"net_kg":27980} // {"topic":"mcp/proxmox","node":"pve-02","cpu":"34%","status":"healthy"} // {"topic":"agent/alert","device":"sensor-17","summary":"reconnected after 4s drop"} // {"topic":"homelab/llm","model":"llama3.1-8b","latency_ms":812} // {"topic":"gate/entry","rfid":"E200341201","status":"granted"} // {"topic":"yard/container","id":"MSCU-2201847","gps":"22.312,70.802"} // {"topic":"weighbridge/tx","gross_kg":42180,"tare_kg":14200,"net_kg":27980} // {"topic":"mcp/proxmox","node":"pve-02","cpu":"34%","status":"healthy"} // {"topic":"agent/alert","device":"sensor-17","summary":"reconnected after 4s drop"} // {"topic":"homelab/llm","model":"llama3.1-8b","latency_ms":812} //

engineering snapshot

Proof, wired into the work.

A compact operating view of the systems I ship and the outcomes they create.

signal / 01

Years in production

Built beyond the demo

Industrial systems, field hardware, backend services, and production debugging.

signal / 02

20+

Systems shipped

From gate to cloud

Device workflows, dashboards, audit trails, and operator tooling in active use.

signal / 03

60-90%

Operational gains

Less manual drag

Faster handling, lookup, reconciliation, monitoring, and alert investigation.

signal / 04

24/7

Realtime visibility

Signals stay live

MQTT, WebSocket, GPS, RFID, event streams, and operator-ready status.

end-to-end ownership

One continuous system—from a noisy field event to a useful human decision.

Explore the evidence

Field

RFID · GPS · sensors

Platform

Spring · events · data

Intelligence

RAG · MCP · agents

Decision

dashboards · alerts

selected work

Systems with a pulse.

Concise system briefs focused on the problem, the engineering decision, and the operational result. Filter by the layer you care about.

showing 6 of 16 systems

Mobile systemopen source

Creator · Android/AI Engineer

Vervan Chat

Mobile AI workflows often send private conversations and documents to hosted services while splitting chat, retrieval, voice, notes, and tools across separate apps.

Keeps chats, documents, embeddings, and inference on-device · combines local chat, document search, voice, notes, tools, and reusable AI workspaces in one app.

KotlinJetpack ComposeMaterial 3LiteRT-LMllama.cppRoomMediaPipe

View source

Mobile systemopen source

Creator · Android Engineer

Gotify+

Self-hosted Gotify users need a modern mobile experience with reliable background delivery, per-application controls, offline access, and faster message discovery.

Brings real-time self-hosted push, per-app notification channels, offline history, search, and resilient background delivery into one modern Android client.

KotlinJetpack ComposeMaterial 3MVVMHiltRetrofitOkHttp WebSocket

View source

Backend systemopen source

Creator · Go/Infrastructure Engineer

Proxmox MCP Server

Infrastructure administration is spread across dashboards and APIs, while AI clients need a controlled way to inspect and operate clusters without making destructive actions the default.

Exposes 27 Proxmox administration tools to MCP clients · keeps destructive operations opt-in while supporting both local and remote integrations.

GoMCPProxmox VE APIAPI Token AuthstdioSSEHTTP

View source

Backend systemopen source

Creator · Go/Infrastructure Engineer

Beszel MCP Server

Checking server health, detailed resource metrics, containers, and alerts requires moving through monitoring views instead of asking one focused operational question.

Turns Beszel server health, resource metrics, container state, and alerts into five focused MCP tools with automatic authentication-token refresh.

GoMCPBeszelPocketBase RESTJWTDockerPodman

View source

AI systemopen source

Edge AI Engineer

rkllama.cpp for RK3576

Running local language models on RK3576 edge boards requires a hardware-specific llama.cpp build path, NPU drivers, and model-format choices that are easy to misconfigure.

Makes local LLM experimentation on RK3576 boards repeatable through an automated NPU-aware build and practical GGUF model support.

C++llama.cppRK3576RKNPU2GGUFQ8_0F16

View source

Backend systemopen source

Creator · Java/MCP Engineer

Advanced File Operations MCP Server

AI-assisted development and operations workflows need more than basic file reads: they also require search, safe transformations, structured-data inspection, checksums, archives, and repeatable batch operations.

Consolidates file management, search, refactoring helpers, structured-data inspection, checksums, archives, and batch workflows behind one MCP interface.

JavaMCPMavenSLF4JJSONCSVZIP

View source

capability stack

I build across the whole signal path.

Open a layer to see the problems I solve, the systems I use, and the value that reaches the operator.

Industrial IoT & Automation SystemsI build systems that connect field devices, operators, backend services, dashboards, and business workflows.

System layer

RFID workflows · ANPR/FastTag integration · GPS/DGPS tracking · MQTT telemetry · Device monitoring · Weighbridge automation · Warehouse automation · Gate workflows · UWB-based tracking · Real-time dashboards · Alerts · Audit logs · Reports

Friction removed

Manual operations, delayed visibility, inconsistent device data, missing audit trails, slow reporting, disconnected workflows, and limited real-time monitoring.

Value delivered

Improved operational visibility · reduced manual workload by up to 60-70% · stronger traceability · real-time alerts · reliable hardware-to-backend data flow.

Java Backend & MicroservicesI design backend systems that are secure, maintainable, observable, and ready for production use.

System layer

Java · Spring Boot · REST APIs · Microservices · Spring Security · OAuth2/OIDC · Keycloak · JWT · API Gateway · RabbitMQ · Redis · PostgreSQL · RBAC · Audit logging · Reporting APIs · Payment workflows · Event-driven systems

Friction removed

Complex workflows, poor service boundaries, weak access control, unreliable processing, difficult debugging, and backend systems that become hard to maintain over time.

Value delivered

20+ Spring Boot services built and maintained · 50% improvement in backend workflow reliability · 30-50% faster production issue diagnosis.

Device Communication & Edge IntegrationI work close to the device layer and understand how data moves from hardware into backend systems.

System layer

Serial communication · UART · SPI · I2C · CAN Bus · MQTT · EMQX · RFID readers · GPS/DGPS modules · Microcontrollers · SBCs · Sensors · Edge devices

Friction removed

Unstable hardware signals, noisy data, device disconnections, inconsistent payloads, communication failures, and unreliable field environments.

Value delivered

More reliable data capture · better device visibility · stronger backend validation · smoother integration between physical devices and software systems.

AI, Agentic AI & MCP SystemsI use AI where it improves visibility, troubleshooting, automation, and decision support.

System layer

Spring AI · Agentic AI · MCP servers · Cloud AI models · Llama · Local LLMs · RAG · Tool calling · AI-assisted monitoring · Human-in-the-loop workflows · AI-based summaries

Friction removed

Teams often have data, logs, dashboards, and alerts, but still spend too much time understanding what happened and what to do next.

Value delivered

AI-assisted monitoring workflows · MCP integrations · natural-language search over operational records · reduced manual infrastructure checks by 60-80%.

Edge AI & Self-Hosted InfrastructureI use my homelab as a practical environment to test infrastructure, local AI, networking, storage, and deployment patterns.

System layer

Raspberry Pi · Radxa · Rockchip NPU devices · RK3588 · RK3576 · Proxmox · Docker · LXC · TrueNAS · OpenWrt · Mac mini M4 · Local LLMs · Monitoring · Self-hosted services

Friction removed

Modern systems need to run beyond a developer laptop — on edge devices, local servers, containers, internal networks, and resource-constrained hardware.

Value delivered

A hands-on test environment for deployment patterns, local AI serving, infrastructure automation, self-hosted monitoring, network design, and edge system experimentation.

Document Intelligence & Data PipelinesI build pipelines that turn unstructured documents and data into structured, searchable, auditable records.

System layer

OCR pipelines · AI-assisted extraction & classification · Manual-review routing · RabbitMQ · Redis · S3-compatible storage · Structured validation · Audit trails · Notification systems · Payment workflow integration

Friction removed

Manual document handling, inconsistent data entry, slow classification, missing validation, and workflows that break down at scale.

Value delivered

3,000+ documents processed monthly · 70% less manual review · full audit tracking across document and notification pipelines.

edge ai & infra

A working lab for edge systems and local AI.

One routed network connects edge services, six Proxmox hosts, shared storage, and local inference hardware—built to test how infrastructure and AI applications behave outside a clean development machine.

lab fabric / live topology

10 endpoints online

uplink

Internet

gateway

OpenWRT router

distribution

Network switch

edge services

Orange Pi Zero

Network entry point for secure access, routing helpers, and service visibility.

Nginx Proxy ManagerVPNBeszelSSO

virtualization / edge cluster

Four Proxmox nodes

Raspberry Pi 5 · PVE 01

8GB RAM · 512GB SSD

Raspberry Pi 5 · PVE 02

8GB RAM · 256GB SSD

Raspberry Pi 5 · PVE 03

8GB RAM · 256GB SSD

Radxa A7A · PVE 04

12GB RAM · 256GB SSD

development / test

Mini PC cluster

Isolated Proxmox capacity for development, application testing, and infrastructure experiments.

Mini PC · DEV 01

16GB RAM · 500GB SSD

Mini PC · DEV 02

16GB RAM · 500GB SSD

storage / backup

CM3588 NAS kit

Shared lab storage

Backups, task data, server volumes, and centralized storage.

local AI fabric

LLM compute

Radxa ROCK 4D

8GB RAM · 256GB SSD · edge inference

Mac mini M4

16GB RAM · LLM runtime & testing

Local inference is exposed to AI applications across the lab when a workload needs it.

hardware / field layer

Raspberry Pi 5 · 8GB · 512GB SSD2x Raspberry Pi 5 · 8GB · 256GB SSDRadxa A7A · 12GB · 256GB SSD2x Mini PC · 16GB · 500GB SSDRadxa ROCK 4D · 8GB · 256GB SSDCM3588 NAS KitOrange Pi ZeroNetwork switchOpenWrt routerMac mini M4 · 16GB

software / control layer

OpenWrt routing and firewallProxmox VE · 6 hostsNginx Proxy ManagerVPN accessBeszel server monitoringSSONAS backups and shared storageLocal LLM inferenceAI application testing

lab outcome

OpenWRT routes one switched fabric across 10 compute and storage endpoints—with proxying, VPN, SSO, monitoring, backups, Proxmox development, and local AI inference in one lab.

field experience

Built in the field. Refined under load.

A career shaped by production constraints, physical devices, real operators, and systems that cannot quietly fail.

current operating context

Suraj Informatics Pvt. Ltd.

IoT Engineer / Backend Engineer / Team Lead

Feb 2022 - Present

I work on backend and IoT systems for industrial automation, enterprise workflows, device tracking, document processing, notifications, access control, payment workflows, and production monitoring. My role combines development, architecture, device integration, deployment support, debugging, and technical leadership.

Designed and maintained Java Spring Boot services for telemetry ingestion, event processing, document workflows, notifications, RBAC, reporting, audit logging, payment workflows, and real-time dashboards.
Worked across backend architecture, API design, database design, device integration, deployment, and production debugging for multiple systems.
Implemented secure authentication and authorization using Spring Security, JWT, OAuth2/OIDC, Keycloak, and custom RBAC.
Built event-driven workflows using RabbitMQ, MQTT, Redis, and WebSocket-based live updates.
Integrated RFID, ANPR, FastTag, GPS/DGPS, serial/UART, sensors, and MQTT-based hardware with backend platforms.
Worked with device communication patterns across UART, SPI, I2C, CAN Bus, MQTT, and industrial hardware integrations.
Led a team of 15+ developers across 10+ project deliveries while contributing hands-on as a developer.

20+ production-grade systems delivered or contributed to200+ Spring Boot services built and maintained50% improvement in backend workflow reliability40-70% efficiency gains across automation workflows30-50% faster production issue diagnosis through better logs, monitoring, and audit trails

Nov 2021

Started Freelance Development

Built a complete business website with backend functionality and payment gateway integration, gaining early experience in full-cycle delivery.

Feb 2022

Joined Suraj Informatics

Started working on backend and IoT systems, building Spring Boot services and integrating real-world devices with software platforms.

2022-2023

Industrial IoT and Backend Systems

Worked on automation systems involving device communication, backend workflows, dashboards, reports, and production deployments.

2023-2024

Lead + Developer Responsibilities

Took ownership of project delivery, backend architecture, team coordination, debugging, deployment support, and production issue resolution.

2024-2025

AI-Enabled Systems and Infrastructure

Expanded focus into Spring AI, MCP servers, RAG, local LLMs, and AI-assisted monitoring while continuing backend and IoT platform work.

Now

Edge AI, Agentic AI, and Practical Automation

Building systems that combine backend platforms, IoT telemetry, infrastructure, local AI, and agentic workflows into useful engineering solutions.

working depth

A stack built around the signal path.

Deepest at the backend and device boundary, with enough product and infrastructure range to own the system around it.

capability telemetrylive profile

Backend / Java95%

IoT / Hardware90%

System Design86%

DevOps / Linux84%

Agentic AI / MCP80%

Frontend62%

how to read this

Broad enough to connect the system. Deep enough to own the hard parts.

The profile reflects daily production work, not a checklist: backend engineering and IoT hardware are the deepest reps; agentic AI and MCP are the fastest-growing edge.

Backend Engineering

JavaSpring BootREST APIsMicroservicesSpring SecurityOAuth2/OIDCKeycloakJWTAPI GatewayRabbitMQRedisPostgreSQLClean ArchitectureHLD/LLDSystem DesignEvent-Driven ArchitectureAudit LoggingProduction Debugging

IoT & Industrial Automation

RFIDANPRFastTagGPSDGPSMQTTEMQXSerial CommunicationUARTSPII2CCAN BusArduinoRaspberry PiESP32ESP8266UWBCustom SensorsGate AutomationWeighbridge AutomationWarehouse AutomationDevice Tracking

AI & Agentic Systems

Spring AIAgentic AIMCP ServersCloud AI APIsLlamaLocal LLMsRAGAI AgentsTool CallingAI-Assisted MonitoringEdge AIModel ServingHuman-in-the-Loop Workflows

Edge & Embedded Systems

RadxaRockchip RK3588Rockchip RK3576RISC-V NPUNPU-Enabled DevicesRaspberry PiSingle-Board ComputersEmbedded LinuxEdge ComputingLocal Inference

DevOps & Infrastructure

DockerLinuxProxmoxLXCAWS EC2NginxTomcatTrueNASOpenWrtSelf-Hosted InfrastructureShell ScriptingMonitoringReverse ProxySSO

Frontend & Full Stack

ReactJavaScriptTypeScriptHTML/CSSAndroid with JavaPHPCodeIgniter

field notes

What the build taught me.

Concise notes from production IoT, backend systems, infrastructure, and applied AI.

notebook / calibrating

The first engineering dispatch is being prepared.

The archive will cover real implementation tradeoffs, field failures, and practical patterns worth reusing.

Open the notebook

Read all posts

get in touch

Let’s Build Practical, Reliable Systems

I am open to engineering roles and collaborations where I can work on backend platforms, IoT systems, Edge AI, Agentic AI, infrastructure automation, and production-ready software that solves real operational problems.

project inquirychannel open

Prefer to talk it through?

Book a call

github.com/anand34577

public activity