Data Science • Machine Learning • Cloud Engineering.
Building production-grade ML systems, not tutorials.
$ engineer --whoami
Name: Kabir Roy
Primary Role: Data Science & Cloud Engineer
Core Focus: Data • Models • Infrastructure • Scalability
Philosophy: Data > Models > Tools🚀 ENGINEERING ROADMAP — 2026 (LOCKED)
Data → Models → Cloud → Deployment
A deliberate engineering trajectory — no noise, no fluff.
Area What I Build
📊 Data Science Pandas • NumPy • EDA • Statistical Modeling
🤖 Machine Learning Scikit-learn • Model Evaluation • Feature Engineering
☁️ AWS Cloud EC2 • IAM • VPC • S3 • CloudWatch
🏗️ Solution Architecture Scalable, fault-tolerant ML system design
⚙️ INFRASTRUCTURE & DEPLOYMENT LAYER
Layer Tools & Practices
🔁 CI/CD GitHub Actions (tested pipelines)
🐳 Containers Docker (local → cloud parity)
🧪 Model Tracking MLflow • Experiment versioning
📊 Observability Logs • Metrics • Monitoring mindset
🧰 ENGINEERING TOOLBOX
| Capability | Description |
|---|---|
| 🔍 Semantic Matching | NLP-based Resume ↔ Job Description scoring |
| 📊 Feature Engineering | Text vectorization & similarity modeling |
| ⚡ Backend Architecture | FastAPI with modular structure |
| 🐳 Deployment Ready | Container-first architecture |
| 🔐 Security Ready | Auth & RBAC-aware design |
Client (Web / API)
│
▼
FastAPI Gateway
│
▼
Processing Layer
├─ Resume Parser
├─ Job Description Analyzer
├─ Semantic Scoring Engine (NLP/ML)
│
▼
Persistence Layer
├─ Metadata Store
├─ Scoring Cache
Designed for horizontal scaling and stateless deployments
☁️ AWS MINI-PROJECTS — HAND-BUILT.
Project What It Proves
🔹 EC2 App Deployment Compute provisioning & SSH security
🔹 IAM Role Design Least-privilege access control
🔹 VPC Networking Subnets • Route Tables • Security Groups
🔹 S3 Storage Object lifecycle & access policies
🔹 CloudWatch Logs, metrics & alerts
Docker Image
│
▼
CI Pipeline (GitHub Actions)
│
▼
AWS EC2 Instance
│
▼
Nginx Reverse Proxy
│
▼
ML Model / FastAPI Application
✔ Local = Cloud parity ✔ Predictable builds ✔ Infra-aware ML system design
Domain Stack
Data Science Python, Pandas, NumPy, Scikit-learn
Cloud AWS EC2, IAM, VPC, S3
MLOps Docker, MLflow, GitHub Actions
Systems Linux, Networking Basics.
📊 ENGINEERING CONSISTENCY
Skill State
AWS Cloud Practitioner Certified ✅
Data Analytics & BI (Power BI) In Progress
Marketing Analytics (NPTEL, IIT KGP) Certified ✅ (93%)
MLOps (Docker, MLflow) Active
❌ ENGINEERING BOUNDARIES (INTENTIONAL)
❌ Buzzword stacking
❌ Fake “Data Science Master” labels
❌ Tutorial-only projects
❌ UI-first engineering
✔ Data-driven architecture ✔ Systems thinking ✔ Depth over hype
🌍 CONNECT:
Consistency beats motivation. Every single time.