Open to opportunities

Aviroop Pal

ML & Backend engineer building scalable AI systems, deep learning models, and production-ready APIs. Focused on NLP, Computer Vision, and privacy-preserving ML.

↗ Get in touch ↓ Resume GitHub LinkedIn x
Experience
AI/ML Intern
BaseApp · Remote, Delhi
2 months
  • Built a real-time Voice-to-Voice AI system with sub-second latency using Faster-Whisper, LLMs, and TTS.
  • Architected WebSocket-based streaming API for bidirectional audio with Redis session management.
  • Ensured production readiness — DoS protection, GPU acceleration, Dockerized deployment.
PythonWebSocketsRedisDockerLLMsOllama
Software Development Intern
Proption Fintech · Remote, Delhi
4 months
  • Developed internal tooling in Go to streamline company operations.
  • Designed financial algorithms including the Supertrend indicator for NIFTY data.
  • Contributed to full-stack feature development and data preprocessing pipelines.
GoBackendFinancial AlgorithmsData Analysis
AI/ML Intern
TherapyU · Hybrid, West Bengal
Jan – Feb 2025
  • Implemented AI/ML algorithms to enhance web application functionality.
  • Collaborated with UI/UX teams for seamless AI feature integration.
  • Handled data analysis, preprocessing, and model optimization for scalability.
PythonMachine LearningAI Integration
AI/ML Intern
NeoCortex LLC · Remote, US
May,2026 - Current
  • Contributed to AI-powered Chrome Extensions focused on automation and productivity workflows.
  • Integrated browser APIs with AI-driven backend systems and REST APIs.
  • Assisted in developing scalable and secure extension architectures for production deployment
JavaScriptChrome Extensions APIAI IntegrationREST APIsGitHub
Projects
Real-Time Voice AI System
Engineered a low-latency conversational Voice-to-Voice AI platform supporting real-time multimodal communication. Built asynchronous APIs and optimized GPU inference.
PythonFastAPIRedisDockerWebSocketsLLMs
NeuroSleepNet (v1.0)
Developed a memory-augmented learning framework to mitigate catastrophic forgetting in AI agents. Implemented latent replay buffers and automatic task-boundary detection mechanisms.
PyTorchHuggingFaceLangChainDockerLLM Systems
Vanta – A Deep Research API
Architected a privacy-first Research-as-a-Service platform automating multi-round research workflows. Built scalable asynchronous infrastructure with job queues and database persistence.
FastAPIPythonPostgreSQLRedisDockerSearXNGLLM APIs
Research Paper Implementations
Implemented transformer architectures from Attention Is All You Need and ResNet models from Deep Residual Learning for Image Recognition. Evaluated models on benchmark datasets.
TransformersPyTorchTensorFlow
Ongoing Research
Federated Learning & Differential Privacy
Researching privacy-preserving federated learning architectures for decentralized AI systems. Exploring differential privacy techniques while building scalable distributed training pipelines.
PyTorchTensorFlowFederated LearningDifferential Privacy
Skills
Languages
PythonGoSQLJavaScriptBash
Deep Learning / ML
PyTorchTensorFlowscikit-learnPandas/NumPySciPy
GenAI / NLP
LLMsLangChainRAGTransformersHugging FaceVector DBs
Computer Vision
OpenCVCNNsResNetYOLOImage Segmentation
Backend Architecture
FastAPIDjangoWebSocketsRedisREST APIs
MLOps / Cloud
DockerKubernetesMLflow / W&BAWS/GCPCI/CD
Education & Certifications
B.Tech — Computer Science & Engineering
Adamas University · 3rd Year · CGPA: 8.0
Generative AI FundamentalsGoogle
Prepare Data for ML APIs on Google CloudDeepLearning.ai
Create and Manage Cloud ResourcesGoogle
Enterprise Design ThinkingIBM
Build a Secure Google Cloud NetworkGoogle
Contact

Let's build something.

Open to ML engineering, backend, or research-adjacent roles. Reach out via email or connect on LinkedIn.

palaviroopoff@gmail.com +91 70445 60675