Job responsibilities

• Experience in Machine Learning Models.
• Design and build LLM-powered solutions for tasks such as summarization, Q&A, chatbots, and semantic search.
• Develop and fine-tune local embedding-based systems for FAQ retrieval, search, and classification using tools like FAISS, Weaviate.
• Build and enhance chatbots with natural language understanding, context tracking, and intelligent fallback mechanisms.
• Implement speech-to-text pipelines using Whisper and similar models for multilingual audio transcription and voice command understanding.
• Apply retrieval-augmented generation (RAG) techniques to combine search with LLMs for more grounded and accurate responses.
• Optimize models through quantization, distillation, or LoRA fine-tuning for performance and efficiency.
• Experience with RNNs, LSTMs, Transformers, and Foundation Models.
• Design and optimize algorithms for entity recognition, text summarization, and semantic search.

Requirements

• 2–3+ years of hands-on experience in data science and NLP research, specializing in low-resource language applications and model optimization.
• Deep expertise in language modeling, embedding techniques, and vector-based semantic search (e.g., FAISS, Annoy), with a strong focus on retrieval-augmented generation (RAG) workflows.
• Solid knowledge of speech recognition and transcription, leveraging models like Whisper, Vosk, and Google Speech-to-Text for real-time and batch processing tasks.
• Experience in chatbot development, including real-time interaction, fallback and context handling, and integration with APIs and platforms such as Telegram and WhatsApp.

Interested candidates can apply by clicking the link provided in the "Apply" button.