Navneeth Premanand

Open for new projectsDubai, UAE · GCC and India

I build the systemssmall businesses run on.

ERPs, CRMs, dashboards, the data pipelines underneath them and the AI features on top. From the first requirements call to the deployed product, I sit in the room with the business and in the codebase.

Start a projectSee the work
Years shipping production software
4 years
shipping production software
Systems in production
5 systems
in production, built end to end
Published research
IEEE
published research on image classification

01Selected work

Three systems, end to end.

Each one started as a conversation about a business problem and ended as software people use every day.

01

ERP for a coaching center

SAT preparation center · 2025

SATisfactory

Problem
Student credits, class schedules, attendance and invoices lived in separate spreadsheets. Credits leaked, classes double-booked, and nobody could say what a student had left.
Result
One system now runs the center from first enquiry to final invoice, built to hold 1000+ students.

NestJS · React · PostgreSQL · Prisma · Redis · BullMQ

Live site

02

Website plus back office

Trading company, Dubai · 2024 to 2025

Rynova Trade

Problem
The company needed a public site that could sell, and a private system that could run procurement, sales, invoicing and reporting. Two vendors would have meant two systems that never agreed.
Result
A single platform: the catalogue and blog the public sees, and the admin dashboard the team runs the business from. Deployed on AWS with automated releases.

React · Node.js · Express · MongoDB · TypeScript · AWS

Live site

03

Process mining tool

Own product · 2024

Minedful

Problem
Process analysts who want to see how work actually flows through a system either pay for enterprise BPM licences or write scripts by hand.
Result
Upload an event log, get a process model, KPIs and a rework analysis in one screen. Self-hostable, no licence.

React · Django · PM4Py · MongoDB · AWS S3 · Graphviz

Live site

Also

  • Fleet analytics dashboardData

    Flo Mobility. Live operations view for 50+ autonomous robots. Cut report latency by about a third by reworking the MongoDB pipelines.

  • ERP and HRMS migrationSystems

    ACZ Global. Moved 40 people from paper to a MERN system on AWS. I gathered the requirements, designed the schema, built the APIs and ran the UAT.

  • Stellar image classificationAI and ML

    Compared VGG16, a parallel CNN and ResNet on star and galaxy images. Published in IEEE.

    Paper

  • Medical imaging pipelineAI and ML

    CNN classification for breast cancer imaging, from exploratory analysis to stakeholder visuals.

02Services

What I can do for you

Six things clients hire me for. Most projects are a mix of two or three.

  1. 01

    Custom ERP and CRM

    Replace the spreadsheets, WhatsApp threads and paper files with one system that knows who your customers are, what they paid for and what they have left.

    NestJS or Node, PostgreSQL, React

  2. 02

    Web platforms with a back office

    A public site that sells, and the admin side the team runs the business from. One codebase, one deploy, no handoff between vendors.

    React, Node, TypeScript, AWS

  3. 03

    Dashboards and reporting

    Put the numbers people ask for every Monday in front of them without anyone exporting a CSV. Built on your existing data, wherever it lives.

    React, SQL, MongoDB pipelines, Power BI or Tableau where it fits

  4. 04

    Data pipelines and analytics

    Get the data out of the systems it is stuck in, clean it, and land it somewhere you can ask questions of it. From robot telemetry to sales ledgers.

    Python, Pandas, SQL, MongoDB aggregation, scheduled jobs

  5. 05

    AI and machine learning features

    Classification, prediction and document understanding built into the tools you already use, with a clear answer on whether the model is actually earning its keep.

    Python, TensorFlow and Keras, scikit-learn, LLM APIs

  6. 06

    Requirements and process mapping

    Sit with the team, map how work actually moves, and write the spec before anyone writes code. Sometimes the answer is a smaller system than you thought.

    Interviews, process maps, SRS documents, user stories

03Capabilities

Three layers, one person.

The application, the data underneath it, and the models on top. I work across all three, which is why the seams between them do not leak.

Build

The application layer. Systems people log into every day.

  • TypeScript, React, Next.js
  • NestJS, Node, Express, Django
  • PostgreSQL, Prisma, MongoDB, MySQL
  • Redis, BullMQ, JWT auth, RBAC
  • Docker, AWS (EC2, S3, Lambda), CI/CD

Data

Getting the numbers out, cleaned, and in front of people.

  • Python, Pandas, NumPy, SQL
  • MongoDB aggregation pipelines
  • Statistical analysis and data modelling
  • Tableau, Power BI, Chart.js, Recharts
  • Process mining with PM4Py

AI and ML

Models where they pay for themselves, and honest reports where they do not.

  • TensorFlow, Keras, scikit-learn
  • CNN architectures for image classification
  • Feature engineering and model evaluation
  • LLM-assisted workflows in production code
  • IEEE-published research on stellar image classification

Not sure which of these you need? Send me the problem. I will tell you what I would build, and whether you need it at all.

Describe your problem

Free, no obligation.

04How I work

Talk first, then build.

  1. 01

    Talk first

    I cannot know what problem I am solving until I have talked to the people who have it. The first week is interviews and process maps, not code.

  2. 02

    Choose the simple architecture

    A well-structured monolith beats microservices you do not need. I pick the design that is easiest to change and debug at your size, not the one that looks best on a slide.

  3. 03

    Build fast, review strictly

    AI tools write the boilerplate, tests and docs. I write the business logic and review everything that ships. That split is how one person delivers what used to take a team.

  4. 04

    Ship and stay

    Deploy, monitor, fix what breaks, and keep iterating after launch. The system is done when the team stops thinking about it.

05About

A bit about me.

I grew up in Bengaluru and spent 24 years there: school at Kendriya Vidyalaya, two degrees at Christ University, and my first two jobs. In 2024 I moved to Dubai with my family.

My work has moved between full-stack development, business analysis, data engineering and machine learning. That was not indecision. I like understanding the whole picture, and the jobs I do best are the ones where I can talk to the client in the morning and change the schema in the afternoon.

That is what I mean by technofunctional. Not a developer who waits for a spec, and not an analyst who hands one over. Both, in one chair.

Navneeth Premanand
I want to be in the room with the client and in the codebase.

  • On the road

    A Royal Enfield Interceptor 650 through Ooty, Mysore, Coorg and Wayanad. The Western Ghats are my reset button.

  • Off the field

    Football when I can, gaming when I cannot.

  • In the kitchen

    I cook and experiment. The kitchen is another lab.

  • After hours

    Self-hosted alternatives, custom ROMs, open-source tools. I like knowing how the things I use actually work.

  • Equivalent exchange

    "To obtain, something of equal value must be lost." Fullmetal Alchemist, and also how I think about trade-offs in software.

06Background

Where I have worked and studied.

  1. May to Nov 2025

    Full Stack Engineer

    Flo Mobility, Bengaluru, remote from Dubai

    Data pipeline and live analytics dashboard for a fleet of 50+ autonomous robots.

  2. Aug 2024 to Apr 2025

    Full Stack Developer, contract

    Rynova Trade LLC, Dubai

    Sole developer on the public site and admin platform, from requirements to AWS deployment.

  3. Feb to Apr 2024

    Business Process Automation Intern

    ACZ Global, Bengaluru

    Mapped business processes and wrote SRS documents and workflow diagrams.

  4. Jun 2023

    AI Engineering Intern

    Open Weaver, Bengaluru

    Benchmarked CNN architectures on medical imaging data.

  5. Sep 2021 to Sep 2022

    Full Stack Developer

    ACZ Global, Bengaluru

    Built the company's CRM and ERP from scratch and led the move off paper for 40 users.

Degrees

  • MSc Data Science

    Christ University, Bengaluru, 2022 to 2024

  • BSc Computer Science, Mathematics and Statistics

    Christ University, Bengaluru, 2018 to 2021

Certifications

  • Google Data Analytics Professional Certificate
  • Business Analysis and Process Management, Coursera
  • Data Science 101, IBM
  • Cybersecurity and IoT, University System of Georgia
  • Cyber Law, Christ University

07Before you write

Questions clients ask first.

How do we start?

A call, usually 30 to 45 minutes, where you tell me how the work runs today and where it hurts. I come back with a written note: what I would build, in what order, and what I would leave out. You decide from there.

We already have a system. Can you work with it?

Yes. Most projects start from something: a spreadsheet, an old CRM, an accounting tool. I map what is there, keep what works, and migrate the data rather than asking anyone to retype it.

Do you work remotely?

Mostly. I am in Dubai and travel to Bengaluru often, so in-person sessions in either city are easy. Everything else runs over calls and a shared workspace.

Who owns the code?

You do. The code, the repository, the hosting accounts and the data all sit in your name. If you ever want someone else to take over, they can.

What happens after launch?

I stay on. Launch is when real usage starts and the small things show up. I fix them, watch the monitoring, and keep iterating for as long as you want me to.

08Contact

Tell me about your project.

Email is fastest. I reply within a day, usually sooner.

Where
Dubai. Remote works, and I travel to Bengaluru often.