Open to AI/ML, Data Science and Flutter + GenAI roles

Portfolio · 2026

I train models that read the body's signals, and build the mobile apps they live in.

First author of a graph neural network paper on seizure detection from IIT Delhi. Flutter developer at Cipher Craft, with four apps live for real clients.

See selected workDownload CV
Based in
Rajkot · --:-- IST
Research
IIT Delhi, IJERCSE ’26
Shipping
Flutter at Cipher Craft
Seizure watch

Channel 01

Live
Graph · 0 nodes · 0 edges
Window clearGCN + SVM · monitoring
99.25%Accuracy
100%Sensitivity

Concept screen: my research model, running in the kind of app I ship.

Selected work

One research paper, one AI product, three apps in production.

Each project opens into a full case study: the problem, the process, the decisions and what I would do next.

A printed EEG trace beside a graph of connected nodes on a desk

01 · Research · Healthcare AI

Reading seizures from a single EEG channel

A Graph Neural Network pipeline that turns one channel of brain activity into a graph, and finds every seizure in the test set. Built at IIT Delhi, published as a first-author paper.

A laptop showing a calm chat interface answering a question

02 · Generative AI · RAG

Ask-My-AI, a recruiter's assistant

A live assistant on artikumari.me that answers recruiters' questions about my work, grounded in my own documents and streamed as it thinks.

Three phones showing a construction management app

03 · Mobile · Flutter · Production

Running construction sites from a phone

A Flutter app for a construction company that replaced paper registers: sites, tasks, materials and people, with each role seeing only what it needs.

Phones showing a mobile CRM pipeline and lead detail

04 · Mobile · Flutter · Production

A CRM that fits in a pocket

Leads, customers and the sales pipeline in one mobile app, so a field team can update a deal in the moment instead of at the end of the week.

A tablet with a trip admin dashboard next to a phone with a customer travel app

05 · Mobile · Flutter · Production

Trip Portal, one product, two apps

A travel product split into an admin app that runs the business and a customer app that sells the trips. Two Flutter apps, released together.

How I work

Practice 01IIT Delhi · PyTorch

In the lab

First-author research on healthcare signals, with models I can read and change.

  • Question the data firstFew labels, uneven classes. Know the constraint before the model.
  • Reframe the problemEEG as a graph, not only a sequence.
  • Write the model, not just the callGCN, CNN and RNN/LSTM coded by hand.
  • Measure what mattersSensitivity first. A missed seizure costs most.

Practice 02Cipher Craft · Flutter

In production

Mobile apps for real clients, built around how each business already works.

  • Learn how the business worksPaper registers and phone calls are the real spec.
  • Design around rolesOwner, engineer, admin, customer. Each sees what it needs.
  • Keep state and APIs cleanBloc/Provider and clear REST contracts.
  • Ship on scheduleSeveral client apps at once, all out on time.
Journey

From a B.Tech in 2022 to research, production and an M.Tech.

Select any bar. Work opens as situation, task, action and result, straight from the CV.

Industry
Research
Study
Learning

2025 – now

Flutter Mobile Application Developer

Cipher Craft Pvt. Ltd.

Mobile engineering · production client delivery

Four live Flutter apps for real clients, built around how each business actually works.

Situation
Construction and CRM clients still used paper and disconnected tools.
Task
Build mobile apps that match how those businesses actually work.
Action
Wrote Flutter/Dart apps with login roles, workflows and REST APIs.
Result
Two live apps in use: Construction Company Management and CRM.
Toolkit

Tools, each tied to where it shipped.

  • 01

    Generative AI

    LLMs, RAG, AI agents, MCP, Prompt engineering, Vercel AI SDK, Gemini API, Embeddings & vector search, Streaming

    Ask-My-AI, live at artikumari.me
  • 02

    Deep learning

    PyTorch, TensorFlow / Keras, GCN, GCN + SVM, RNN / LSTM, CNN, Data augmentation, scikit-learn

    EEG seizure paper, IJERCSE 2026
  • 03

    Data science

    Python, NumPy, Pandas, SQL, EDA, Feature engineering, Evaluation

    Gatim AI internship
  • 04

    Healthcare signals

    Single-channel EEG, Seizure classification, Class imbalance, Graph modelling of biosignals

    IIT Delhi research
  • 05

    Mobile

    Flutter, Dart, Bloc, Provider, REST APIs, Firebase, Role-based access, iOS & Android

    Four live client apps
  • 06

    Web & backend

    React, Next.js, TypeScript, Node.js, C# / .NET, ASP.NET

    artikumari.me