Senior AI/ML Engineer

• Architect and evolve the multi-agent orchestration platform, including plugins, tool-use pipelines, observability hooks, and voice, telephony, and messaging channel adapters
• Design and implement voice AI pipelines using STT, real-time streaming TTS, VAD, and SIP/RTP telephony integration with sub-300 ms latency targets
• Build and maintain RAG pipelines with retrieval quality measurement, re-ranking, and hybrid vector and keyword search
• Define MCP server architecture and tool-use contracts for internal and third-party integrations
• Fine-tune and evaluate LLMs using LoRA, QLoRA, and DPO for customer support, classification, and structured extraction
• Benchmark model quality using automated evaluations, human preference data, and speech/RAG-specific metrics
• Manage model lifecycle, experiment tracking, versioning, reproducibility, and production promotion
• Own AI observability with Langfuse tracing, LLM call instrumentation, cost tracking, and quality regression alerting
• Define and enforce guardrails for hallucination detection, PII redaction, output safety scanning, and rate limiting
• Build data ingestion, preprocessing, feature, and continual-learning pipelines
• Drive ML CI/CD, including automated evaluation gates, shadow deployments, canary releases, and rollback triggers
• Set AI architectural standards, conduct design reviews, and own ADRs
• Mentor ML engineers and applied scientists
• Collaborate with Product and Commercial teams to translate business problems into ML formulations and success metrics
• Engage with external research partners and monitor emerging research and open-source developments

• 8+ years in ML Engineering, Applied AI, or Research Engineering
• At least 2 years in a lead or staff-level role
• Deep, hands-on production experience with LLMs, including fine-tuning, RLHF/DPO, prompt engineering, RAG, and tool use
• Fluent in Python and experienced with PyTorch, HuggingFace Transformers, and PEFT/LoRA
• Experience with LLM inference serving using vLLM, TensorRT-LLM, or TGI in a latency-sensitive production environment
• Practical knowledge of agentic frameworks, multi-agent coordination, tool-call orchestration, context/memory management, and observability
• Solid understanding of MLOps, including MLflow/W&B, Docker/Kubernetes, model registries, and CI/CD for ML
• Awareness of LLM-specific risks and production mitigations, including hallucination, prompt injection, data leakage, fairness, and privacy
• Strong communication skills for design documentation, architecture reviews, and explaining tradeoffs to non-technical stakeholders
• Proof of eligibility to work in the country applied for
• Experience with speech AI or real-time audio systems is a strong plus
• Experience with voice pipelines, multi-hop RAG, MCP, open-source ML contributions, multimodal models, or quantization techniques is nice to have

• No relocation packages or visa sponsorship
• Diversity and inclusion commitment
• ESG and sustainability initiatives