I will build and deploy a custom machine learning model with fastapi

S
spandan_ml
S
spandan_ml
Spandan Das
build and deploy a custom machine learning model with fastapi
Full Screen
build and deploy a custom machine learning model with fastapi
Full Screen
build and deploy a custom machine learning model with fastapibuild and deploy a custom machine learning model with fastapi

About this gig

Are you looking for a custom ML model that actually works in production not just a Jupyter notebook?

I build end-to-end machine learning solutions: from data preprocessing and model training to deployment as a live REST API. My work spans image classification, NLP, fraud detection, time series forecasting, and reinforcement learning.

What you get:

  • Trained model on your dataset
  • FastAPI backend with documented endpoints
  • Streamlit dashboard (optional)
  • GitHub repo with clean, commented code
  • Deployment to HuggingFace Spaces or your preferred platform

My recent projects include a real-time fraud detection system (97.6% AUC), a medical imaging platform with 4 CNNs, and a financial sentiment engine with live data pipelines.

I communicate entirely via text fast replies, clear updates, no surprises.

  • Application Type

    • Web Application

  • Desktop Frameworks

    • Other

  • AI Type

    • Other

  • Programming Language

    • Python

  • Web Frameworks

    • Other

  • No & Low-Code Builders

    • Replit

Get to know Spandan Das

Spandan Das

ML Engineer, Computer Vision, NLP and Model Deployment

  • FromIndia
  • Member sinceApr 2026

  • Languages

    English
I build production-grade ML systems — from model training to deployed APIs. My work spans computer vision (CNNs, GRAD-CAM), NLP (BERT, sentiment analysis), reinforcement learning, and graph neural networks, shipped with FastAPI and Streamlit. Recent projects: real-time fraud detection at 97.6% AUC (GraphSAGE + Kafka), multi-modal medical imaging (4 ResNet50 CNNs + DQN triage), and a financial sentiment engine (FinBERT + BiLSTM + XGBoost) with live data pipelines.

My Portfolio

Related tags