3 Whisper Jobs

We're looking for an experienced AI/ML Technical Lead to architect and drive the development of our intelligent conversation engine. Youll lead model selection, integration, training workflows (RAG/fine-tuning), and scalable deployment of natural language and voice AI components. This is a foundational hire for a technically ambitious platform. Key Responsibilities AI System Architecture: Design the architecture of the AI-powered agent including LLM-based conversation workflows, voice bots, and follow-up orchestration. Model Integration & Prompt Engineering: Leverage APIs from OpenAI, Anthropic, or deploy open models (e.g., LLaMA 3, Mistral). Implement effective prompt strategies and retrieval-augmented generation (RAG) pipelines for contextual responses. Data Pipelines & Knowledge Management: Build secure data pipelines to ingest, embed, and serve tenant-specific knowledge bases (FAQs, scripts, product docs) using vector databases (e.g., Pinecone, Weaviate). Voice & Text Interfaces: Implement and optimize multimodal agents (text + voice) using ASR (e.g., Whisper), TTS (e.g., Polly), and NLP for automated qualification and call handling. Conversational Flow Orchestration: Design dynamic, stateful conversations that can take actions (e.g., book meetings, update CRM records) using tools like LangChain, Temporal, or n8n. Platform Scalability: Ensure models and agent workflows scale across tenants with strong data isolation, caching, and secure API access. Lead a Cross-Functional Team: Collaborate with backend, frontend, and DevOps engineers to ship intelligent, production-ready features. Monitoring & Feedback Loops: Define and monitor conversation analytics (drop-offs, booking rates, escalation triggers), and create pipelines to improve AI quality continuously. Qualifications Must-Haves: 5+ years of experience in ML/AI, with at least 2 years leading conversational AI or LLM projects. Strong background in NLP, dialog systems, or voice AI preferably with production experience. Experience with OpenAI, or open-source LLMs (e.g. LLaMA, Mistral, Falcon) and orchestration tools (LangChain, etc.). Proficiency with Python and ML frameworks (Hugging Face, PyTorch, TensorFlow). Experience deploying RAG pipelines, vector DBs (e.g. Pinecone, Weaviate), and managing LLM-agent logic. Familiarity with voice processing (ASR, TTS, IVR design). Solid understanding of API-based integration and microservices. Deep care for data privacy, multi-tenancy security, and ethical AI practices. Nice-to-Haves: Experience with CRM ecosystems (e.g. Salesforce, HubSpot) and how AI agents sync actions to CRMs. Knowledge of sales pipelines and marketing automation tools. Exposure to calendar integrations (Google Calendar API, Microsoft Graph). Knowledge of Twilio APIs (SMS, Voice, WhatsApp) and channel orchestration logic. Familiarity with Docker, Kubernetes, CI/CD, and scalable cloud infrastructure (AWS/GCP/Azure). What We Offer Founding team role with strong ownership and autonomy Opportunity to shape the future of AI-powered sales Flexible work environment Competitive salary Access to cutting-edge AI tools and training resources Post your resume and any relevant project links (GitHub, blog, portfolio) to career@sourcedeskglobal.com. Include a short note on your most interesting AI project or voicebot/conversational AI experience.

What would you do? System Design: Architect and design end-to-end speech processing pipelines, from data acquisition to model deployment. Ensure systems are scalable, efficient, and maintainable. Advanced Modeling: Develop and implement advanced machine learning models for speech recognition, speaker diarization, and related tasks. Utilize state-of-the-art techniques such as deep learning, transfer learning, and ensemble methods. Research and Development: Conduct research to explore new methodologies and tools in the field of speech processing. Publish findings and present at industry conferences. Performance Optimization: Continuously monitor and optimize system performance, focusing on accuracy, latency, and resource utilization. Collaboration: Work closely with product management, data science, and software engineering teams to define project requirements and deliver innovative solutions. Customer Interaction: Engage with customers to understand their needs and provide tailored speech solutions. Assist in troubleshooting and optimizing deployed systems. Documentation and Standards: Establish and enforce best practices for code quality, documentation, and model management within the team. Required Skills 2+ years of experience in speech processing, machine learning, and model deployment. Demonstrated expertise in leading projects and teams. Technical skills: • Excellent knowledge in Python / Java programming. • In-depth knowledge of speech processing frameworks like, Wave2vec, Kaldi, HTK, DeepSpeech and Whisper. • Experience with NLP, STT, Speech to Speech LLMs and frameworks like Nvidia NEMO, PyAnnote. • Proficiency in Python and machine learning libraries such as TensorFlow, PyTorch, or Keras. • Experience with large-scale ASR systems, speaker recognition, and diarization algorithms. • Strong understanding of neural networks, sequence-to-sequence models, transformers and attention mechanisms. • Familiarity with NLP techniques and their integration with speech systems. • Expertise in deploying models on cloud platforms and optimizing for real-time applications. Good to have: • Experience with low-latency streaming ASR systems. • Knowledge of speech synthesis, STT (Speech-to-Text) and TTS (Text-to-Speech) systems. • Experience in multilingual and low-resource speech processing.

Role & responsibilities Build Audio ML/DL models for given requirement Build models from scratch wherever required Must be able to code relevant research papers independently Tune to improve accuracy and latency Support to satisfy the deployment requirements of the product Apply transfer learning wherever required Training, fine tuning and optimization of different flavors of transformer models Hands on experience with LLM, fine tuning and Optimization of LLM Audio AI Modelling and tuning experience Audio AI concepts, ML/DL working experience in Pytorch/Tensorflow/Keras framework Variations of CNN, RNN, Attention Mechanism, Transformers and large models like wav2vec, whisper Proficiency in machine learning/audio pre-processing libraries/frameworks such as Librosa, Kaldi, sklearn, python speech features, Numpy, Scipy, Pandas, Matplotlib. Signal Processing fundamentals, understanding the data characteristics and wave signals Transfer Learning Experience in Acoustic models for speech recognition systems. Experience in Voice activity detection with noise analysis & denoising Experience in WakeupWord, ASR Working experience of LLM fine tuning, optimization and performance improvement. Audio AI Audio related certification Tools for creating Synthetic data Experience working in Edge devices Experience in noise augmentation and mixing techniques