← Back to the doors

AI/ML Engineering

Intelligent machine learning models, neural networks, LLM agents and predictive pipelines engineered for high accuracy, efficiency and production deployment.

AI Agent · LLM & RAG

Autonomous RAG Agent

Context-aware enterprise knowledge assistant

An intelligent multi-agent pipeline combining retrieval-augmented generation (RAG) with local vector embeddings. Queries enterprise documents with low latency, cites verified sources, and self-evaluates hallucination rates in real time.

  • LangChain
  • LlamaIndex
  • PyTorch
  • Vector DB
AI agent interaction interface

Computer Vision · Deep Learning

Neural Vision AI

Real-time multi-object detection and tracking

A high-speed convolutional neural network fine-tuned for edge devices. Detects, segments and tracks objects across live video streams at 60+ FPS with sub-millisecond inference times and minimal memory footprint.

  • PyTorch
  • YOLOv8
  • OpenCV
  • TensorRT
Real-time vision pipeline tracking

Machine Learning · Anomaly Detection

Fraud Detection Engine

Real-time transactional anomaly classification

An ensemble machine learning model combining gradient boosting and isolation forests to flag anomalous financial transactions with 99.4% ROC-AUC. Built with automated drift detection and continuous retraining pipelines.

  • XGBoost
  • Scikit-Learn
  • MLOps
  • FastAPI
Anomaly scoring monitor

Speech AI · Audio Processing

Voice Sentiment Classifier

Acoustic emotion & intent recognition

End-to-end transformer model processing raw audio spectrograms to classify speaker sentiment, pitch cadence and semantic intent with multi-lingual audio adaptation.

  • Transformers
  • HuggingFace
  • PyTorch
  • Librosa
Spectrogram & sentiment analysis