Jayan Subramanian

AI Engineer, India

I create tools and systems shaped by clean thinking and solid engineering.

I approach engineering with a focus on clarity, structure, and long-term reliability. I enjoy turning complex or ambiguous problems into systems that feel clean, intuitive, and scalable. My work emphasizes thoughtful design, precise reasoning, and solutions that stay maintainable as they grow.

View résumé View GitHub

About

Why I build structured, end-to-end projects

I’m a B.Tech Artificial Intelligence Engineering student at Amrita Vishwa Vidyapeetham who combines ML fundamentals with practical software engineering to build systems that are both explainable and extendable.

Who I am

AI student who turns ideas into grounded, reproducible workflows.

• Currently completing my AI Engineering degree with a strong foundation in ML, CV, NLP, data analysis, and systems engineering.

• I build practical tools—analysis utilities, LLM-driven agents, dashboards, UI replicas, and ML experiments—implemented using Python, FastAPI, and React/Next.js.

• My GitHub showcases consistent hands-on work across models, APIs, and front-end engineering, documented for clarity and reproducibility.

What sets me apart

Clear reasoning, maintainable code, and strong fundamentals.

• I prioritize correctness and clarity—verifying assumptions, structuring code cleanly, and ensuring each component is easy to test and extend.

• Comfortable across the stack, from ML pipelines and data preprocessing to API layers and polished front-end interfaces.

• I maintain detailed repo structures, environment notes, and walkthroughs so collaborators can understand and build upon my work immediately.

Skills

Stacks I reach for

Grouped to mirror how I ship: applied AI research, product engineering, and operational rigor. Logos display when official artwork is available; otherwise I keep a neutral icon placeholder.

Programming Languages

Python

C++

JavaScript/TypeScript

SQL

Bash

PowerShell

AI & ML

PyTorch

TensorFlow/Keras

Scikit-Learn

Hugging Face

OpenCV

LangChain

Pinecone

MLflow

Web Development

FastAPI

Node.js

Express

React

Next.js

Vue.js

Hugo

tRPC

Cloud & DevOps

Docker

OCI

GCP

Git

GitHub

Tools & Testing

Postman

Selenium

BeautifulSoup

Bun

Deno

Featured

Projects with a clear arc

Four deep dives that show motivation, the systems thinking applied, and why the solution resonated.

AI-first study companion for document comprehension

DocuMate

Students often struggle to understand lengthy lecture notes, textbooks, and reference PDFs, spending most of their time extracting definitions, key ideas, and exam-relevant topics instead of actually learning.

Approach

Designed an upload-to-study pipeline that breaks documents into structured segments, identifies important concepts, generates clear explanations, and produces practice questions grounded in the original material.

Impact

Students could revise entire chapters far more quickly, with structured summaries and question banks helping them prepare confidently without manual note-taking.

Role

AI + full-stack: document segmentation, retrieval pipelines, explanation and question-generation modules, backend APIs, and an interactive interface for reviewing study-ready outputs.

FastAPICrewAINext.jsChromaDBLangchainHuggingFaceAppwrite

Open DocuMate demo

Discovery sandbox for PC releases

GameSiphon

Gamers needed a way to find PC titles that match their preferences—genre, style, difficulty, system requirements—but existing platforms made filtering and comparison slow or limited.

Approach

Built a full game-discovery catalog with precise filters, fast search, sortable attributes, and clean detail views so users can quickly narrow down games based on interests and hardware.

Impact

Players could identify suitable games far more efficiently, generating personalised shortlists without scrolling through dozens of tabs.

Role

Full-stack + data modeling: schema design, optimized query and filter logic, API integration, and UI architecture for browsing, inspecting, and comparing game metadata.

Next.jsSupabasePostgreSQLtRPC

Open GameSiphon demo

Event analysis tool for causal narrative discovery

Plotline

Understanding why an event happened often requires checking multiple sources, timelines, and statements, making it difficult to form a clear and defensible explanation.

Approach

Developed a system that extracts events, identifies potential causes, links evidence, and organizes everything into a synchronized timeline that helps users build causal narratives step by step.

Impact

Users could move from scattered sources to a coherent, evidence-backed explanation in minutes, supported by a structured timeline and grounded causal links.

Role

AI systems + backend + UX: causal schema design, event-to-evidence linking logic, agent-based analysis endpoints, and a responsive timeline interface for exploring hypotheses.

Next.jsTypeScriptFastAPICrewAIStarlette

Open Plotline demo

Precision oncology console for pathway-aligned care

Pathway Analysis

Clinical teams had to reconcile biomarkers, staging guidelines, and available resources manually before preparing oncology recommendations, slowing case preparation.

Approach

Built a coordinated workspace that aligns patient details, molecular indicators, and guideline-based steps into a single structured flow.

Impact

Care teams could generate clear, actionable case summaries much faster, with all decision-relevant information presented in context.

Role

Full-stack + clinical data engineering: platform architecture, staging and biomarker workflows, interface design, and integration of a collaboratively developed GAT model for pathway interpretation.

Next.jsTypeScriptFastAPIPyTorch (GAT)Firebase

Open Pathway demo

Other Work

Additional explorations

Smaller surfaces, ops tooling, and writing that inform how I build and communicate.

SmartCity Ops Console

Telemetry, nutrition ops, and imaging AI snapshots unified into a single operator dashboard.

ObservabilityAI Ops

Dive deeper →

n8n Automation Dashboard

Kanban swimlanes, GTM news pulse, and inbox briefs for automation squads.

AutomationDashboards

Dive deeper →

Learning-rate essays

Long-form articles documenting insights and discoveries from my learning journey.

WritingReflection

Start reading →

Live Sandboxes

Click-through replicas

Each card opens a fully navigable route powered by curated mock data. Toggle your theme and the visuals update in place.

DocuMate

AI-assisted study companion for concept mining, question drills, and study guide synthesis.

Open demo

GameSiphon

Game discovery sandbox with real catalog data, filters, and capsule reviews.

Open demo

Plotline

Causal analysis explorer scoped to the 2008 financial crisis with sourced evidence.

Open demo

Pathway Analysis

Clinician console for multi-omic risk review and pathway-aligned protocols.

Open demo

SmartCity

Operations console spanning pipeline telemetry, NutriBot recommendations, and imaging AI.

Open demo

n8n Dashboard

Automation ops cockpit with kanban lanes, news pulse, and inbox briefs.

Open demo

Education

Academic grounding & recognitions

My coursework and project experience span machine learning, algorithms, software engineering, and data-intensive systems.

Amrita Vishwa Vidyapeetham

B.Tech · Artificial Intelligence Engineering

• Core subjects include Machine Learning, Deep Learning, NLP, Computer Vision, Big Data Analytics, DBMS, Networks, and Software Engineering.

• Developed strong fundamentals in modeling, optimization, probability, and full-stack implementation.

• Consistent academic performance with hands-on project work across multiple AI and engineering domains.

Highlights

• Built a diverse portfolio of ML, DL, and web-engineering projects available publicly on GitHub.

• Completed multiple independent projects exploring LLM workflows, causal reasoning, and structured analysis tools.

• Active on LinkedIn and GitHub documenting learning progress, experiments, and build logs.

• Updated blog posts and articles on my learning journey.

Résumé & Code

Dig deeper

One pager for quick context, repo references for every build, and a direct line if you’d like to collaborate.

Résumé

An overview of my experience, technical depth, and delivery track record.

View résumé PDF

Repos

Code for every demo lives in the open—pattern libraries, API stubs, and evaluation notebooks.

Explore repos

Contact

Let's build the next narrative together.

I’m selective about engagements so each partnership gets the right depth. Share a brief, a problem statement, or even just a hunch—I'll reply with a POV and a path forward.

Email me GitHub