AS
Open to Data Science roles

Hi, I'm Alok — I build
ML systems that ship.

Entry-level data scientist working across Python, SQL, computer vision, and applied ML — from real-time object detection to local RAG pipelines. I like turning messy data into things people actually use.

5
PROJECTS SHIPPED
6+
ML TOOLS USED
2025
BCA GRADUATE
100%
DEPLOYED, NOT JUST NOTEBOOKS
About me

From raw data to real decisions.

I'm a Bangalore-based data scientist with a BCA background and hands-on experience across the full ML lifecycle — data cleaning, feature engineering, model development, and deployment.

My favorite part isn't the model, it's what happens after: getting it in front of real users. I've shipped a YOLO-based computer vision system, a local RAG pipeline for industrial part search, and several deployed analytics dashboards.

I'm looking for a team where I can keep doing exactly that — solving real business problems with data, not just building notebooks that never leave my laptop.

Machine Learning

Regression, classification, clustering, CNNs

Computer Vision

YOLO, OpenCV, real-time detection

Analytics

Power BI, Streamlit dashboards

Deployment

Streamlit, MLflow, GitHub, FAISS

Skills

My toolkit

The languages, libraries, and tools I reach for when building something end-to-end.

Languages
PythonSQL
ML / Deep Learning
RegressionClassificationClusteringCNNYOLOFeature Eng.
Libraries
PandasNumPyScikit-learnTensorFlowPyTorchOpenCV
Visualization
Power BISeabornMatplotlib
Deployment
StreamlitGitHubMLflowFAISS
Databases
MySQL
PythonDaily use
Machine LearningCore focus
Computer VisionProject-tested
SQL & DashboardsApplied
Projects

Things I've built

Five end-to-end projects — each one taken from raw data through to a deployed, usable interface.

Smart Vision AI

Real-time object detection and classification system with a full deployment pipeline.

85% mAP~45ms inference
  • Trained on 1,200 labeled images across 6 object classes
  • Modular pipeline: preprocessing → inference → visualization
  • Live Streamlit demo interface
PythonYOLOOpenCVStreamlit

PatrolIQ

Urban safety intelligence dashboard for crime-hotspot identification.

9K records analyzed7 hotspots found
  • K-Means clustering + PCA for pattern discovery
  • Interactive dashboards for time & location trends
PythonClusteringMLflowStreamlit

Business Performance Monitor

Automated KPI tracking dashboard replacing manual spreadsheet reporting.

12 KPIs tracked
  • Real-time alerts on revenue & acquisition metrics
  • Monthly performance-vs-target views
PythonPandasPower BI

EMI Prediction System

Regression system predicting monthly EMI from loan parameters.

R² = 0.91
  • Benchmarked Linear Regression, Random Forest, Gradient Boosting
  • Deployed as a real-time Streamlit prediction tool
Scikit-learnPandasStreamlit
Flagship project

Nexus — Local RAG for Industrial Supply Chains

Semantic search and substitution recommendations for industrial parts, running fully local with no external API calls.

~500 parts indexed<1s search time
  • FAISS + Hugging Face embeddings for semantic similarity search
  • Llama 3 (via Ollama) explains substitution reasoning in plain language
  • Interactive Streamlit app for search + recommendations
FAISSLlama 3OllamaStreamlit
Background

Education & training

Bachelor of Computer Applications (BCA)

Completed 2025

Data Science Training Program

GUVI HCL
  • End-to-end ML workflow: EDA, model building, evaluation
  • Applied to real datasets for classification and regression
Contact

Let's talk about your team.

Open to data science, ML, and analytics roles — happy to walk through any of these projects in more depth.