Native apps
Focus tools, puzzles, and finance utilities designed in Swift and SwiftUI — calm interfaces, real shipping constraints.
Rohit
App & AI Developer
App & AI Developer · Canada
Building appsand small minds.
I ship iOS apps
Software engineer and entrepreneur building native apps, compact language models, and the interfaces that sit between them.
TypeScriptPythonSwiftSwiftUINode.jsNext.jsReactPostgreSQLpgvectorMongoDBRedisPyTorchHugging FaceLangChainFastAPIRAGOllamaOn-device AITypeScriptPythonSwiftSwiftUINode.jsNext.jsReactPostgreSQLpgvectorMongoDBRedisPyTorchHugging FaceLangChainFastAPIRAGOllamaOn-device AI
About
App developer. Model tinkerer.Same person.
I'm Rohit — online I go by rohitoncode. I live in Canada and work across the stack that actually ships: Swift for Apple platforms, TypeScript and Node.js for the web, PostgreSQL when the data has to last, and Python when a model needs to think.
The work I care about sits at the seam of product and intelligence. I design iOS apps end to end, then train small language models and task-specific chatbots — PyTorch, Hugging Face, LangChain, FastAPI — that can run close to the user instead of sending every sentence to a distant API.
When I'm not shipping, I write about agents, SwiftUI, and local inference, and share how the apps come together on YouTube and Instagram.
LLMs & chatbots
PyTorch, Hugging Face, LangChain, and FastAPI for compact models, RAG, and conversational UIs I can fine-tune and actually run.
Full-stack product
Node.js, Next.js, TypeScript, and PostgreSQL — with MongoDB when the document shape fits — so an app has a real backend, not only a face on the phone.
Languages I ship in
The stack behind the products
95%
TypeScript
daily language
95%
Python
daily language
85%
Swift
daily language
Industry stack
What production work actually runs on
Node.js services, PostgreSQL with pgvector for retrieval, and a Python LLM path — PyTorch, Hugging Face, LangChain, FastAPI — beside the Swift and TypeScript I ship in the apps.
Languages
TypeScriptPythonSwiftSQL
App & web
SwiftUINext.jsReactNode.js
Data
PostgreSQLpgvectorMongoDBRedis
Python / LLM
PyTorchHugging FaceLangChainFastAPIRAGOllama
Selected work
Apps and experiments
25:00
Focusify
iOSGoals & distraction tracker
A calm iOS companion for setting goals, running focus sessions, and cutting digital noise. Built in Swift with a deliberate, low-friction interface.
SwiftSwiftUI
Download on theApp Store
Coming soon onAndroid
Coming soon onAndroid3
5
7
9
2
Sudoku Premium
iOSPuzzle, refined
A distraction-free Sudoku board with hand-tuned difficulty, tactile animation, and nothing competing for attention except the next number.
SwiftSwiftUI
Download on theApp StoreComing soon onAndroid
Coming soon onAndroidFinsight
iOSFinancial insights
Spending, broken into charts that actually read. Trends, categories, and a layout that makes money feel inspectable instead of noisy.
SwiftSwiftUI
Download on theApp StoreComing soon onAndroid
Coming soon onAndroidSummarize this session and keep it local.
Done. Three takeaways, no cloud round-trip.
Chat Lab
LabSmall models, real conversations
A personal studio for compact LLMs and task-specific chatbots. PyTorch and Hugging Face for the model, LangChain and PostgreSQL (pgvector) for retrieval, FastAPI and Node.js at the edge.
PythonPyTorchLangChainFastAPIPostgreSQLNode.js
Personal research studio
Collaborate
Building an app or a small model?
I like projects where the product and the intelligence have to live in the same room — Swift, Node.js, PostgreSQL, and a Python LLM stack.