Company:
Target Level:
Department:
Role Type:
Location:
About Avaya Core Software & Competency
Avaya is a global leader in enterprise customer experience (CX) and unified communications solutions, powering millions of daily customer interactions across the Fortune 500. Avayas core software platforms deliver mission-critical, high-availability telephony, unified communications (UC), and omnichannel contact center (CC/CCaaS) infrastructures. Our core competencies span carrier-grade SIP signaling, low-latency distributed media processing, secure edge border control, and hybrid cloud orchestration. Building on decades of real-time communication excellence, Avaya is fundamentally infusing state-of-the-art Artificial Intelligence across its software stacktransforming traditional enterprise telephony into highly autonomous, context-aware, and agentic communication fabrics.
Position Overview
AI Systems Engineer (Level 20-IV or above)
Operating at the intersection of production-grade AI/ML engineering and high-performance C/C++/Go systems, you will bridge the gap between cutting-edge AI model serving and carrier-grade RTC infrastructure (SIP, WebRTC, RTP/SRTP) to deliver sub-second, intelligent voice and multimodal interactions at scale.
Role Summary & Technical Profile
Dimension
Key Focus Areas
Core AI Expertise (Primary Focus)
Real-Time Streaming ASR/STT, Low-Latency TTS, Generative AI & Agentic Workflows, Real-Time LLM/SLM Orchestration, Voice Biometrics, Conversational Intelligence, Streaming RAG, Audio DSP for AI
AI Runtime & Frameworks
vLLM, TensorRT-LLM, Triton Inference Server, ONNX Runtime, PyTorch, Hugging Face, LangChain/LlamaIndex, Vector Databases
RTC & Media Infrastructure
SIP (RFC 3261), RTP/RTCP, WebRTC, SDP, STUN/TURN/ICE, TLS 1.3 / mTLS, SRTP, Audio Codecs (Opus, G.711)
Programming Languages
Modern C/C++, Python, Go, Java, Shell scripting
Platform & OS
Linux (systems programming, kernel networking, socket tuning, eBPF), Docker, Kubernetes
Enterprise Domains
Contact Center as a Service (CCaaS), Unified Communications (UC), Customer Experience (CX) Automation
Key Responsibilities
1. AI Architecture & Real-Time Generative Intelligence (Core Focus)
- Lead the architecture and production deployment of low-latency AI inference pipelines directly coupled with live audio streams and enterprise SIP/WebRTC sessions.
- Design and deploy real-time Conversational AI pipelines incorporating streaming Speech-to-Text (ASR), streaming Text-to-Speech (TTS), and streaming LLM token generation to achieve interactive human-like conversational latency (sub-200ms turnaround).
- Architect autonomous, multi-agent AI systems (agentic workflows) for automated call resolution, real-time agent copilot assistance, dynamic intent recognition, and contextual sentiment tracking.
- Implement streaming Retrieval-Augmented Generation (RAG) architectures with enterprise vector search to inject dynamic caller context and corporate knowledge bases into live interactions.
- Optimize, quantize, and serve state-of-the-art open-source and proprietary models (LLMs, SLMs, audio transformers) using high-performance runtimes like TensorRT-LLM, vLLM, and Triton Inference Server.
- Integrate AI-driven audio pre-processing and enhancement pipelines, including Voice Activity Detection (VAD), speaker diarization, background noise reduction, and acoustic echo compensation.
2. Real-Time Communications & Systems Integration
- Design resilient interfaces and microservices connecting high-throughput C/C++ media engines and SIP signaling stacks with asynchronous Python/Go/Java AI model-serving pipelines.
- Ensure secure signaling and media transmission utilizing mTLS, TLS 1.3, and SRTP, upholding strict zero-trust enterprise security requirements.
- Architect NAT traversal and edge connectivity utilizing STUN, TURN, and ICE across enterprise firewalls and distributed edge gateways.
- Conduct deep-dive profiling, latency attribution, and memory management optimizations across Linux-based distributed media and AI microservices.
3. Strategic Technical Leadership & AI Governance
- Act as a principal technical authority on AI systems engineering, guiding architectural roadmaps across UC and CC product lines.
- Establish rigorous benchmarking, evaluation metrics, and guardrails for AI safety, model drift, hallucination detection, and real-time inference latency under peak enterprise call volumes.
- Collaborate with cross-functional architecture and product teams to translate complex AI research into robust, production-ready enterprise software.
Qualifications & Requirements
Required Technical Expertise
Artificial Intelligence & Machine Learning (Primary Requirement):
- Deep, proven experience building and deploying production-grade AI systems, specifically in streaming audio or real-time natural language pipelines.
- Strong hands-on mastery of streaming ASR/TTS architectures (e.g., Whisper, Conformer, FastConformer, or enterprise speech streaming engines).
- Demonstrated experience in LLM application architecture: model fine-tuning, prompt engineering, agentic orchestration frameworks, and low-latency inference serving engines (vLLM, TensorRT-LLM, Triton).
- Strong background in real-time vector search, embedding models, and streaming RAG pipelines.
- Understanding of audio DSP techniques related to ML preprocessing (MFCC, Mel spectrograms, audio resampling, silence/noise removal).
Systems & Software Engineering:
- High proficiency in
Python
(for modern AI/ML development) and C/C++
or Go
(for low-latency systems and media infrastructure), with working knowledge of Java
and robust Shell/Bash
scripting. - Solid foundation in Linux systems programming, POSIX IPC, asynchronous I/O (epoll, io_uring), multithreading, and socket programming.
- Familiarity with debugging tools (gdb, perf, Valgrind) and network packet analyzers (Wireshark, tcpdump).
Real-Time Protocols & Enterprise Networking:
- Hands-on experience or deep architectural familiarity with SIP (RFC 3261), SDP, RTP/RTCP, and WebRTC.
- Solid understanding of NAT traversal protocols (STUN, TURN, ICE) and secure transport protocols (TLS/mTLS, SRTP).
Preferred Domain Experience
- Demonstrated domain experience in Unified Communications (UC) or Contact Center (CC / CCaaS) platforms.
- Experience integrating AI services into telephony/media frameworks, softswitches, or SBCs (e.g., FreeSWITCH, Asterisk, Kamailio, mediasoup, Janus).
- Hands-on experience deploying scalable containerized AI services on Kubernetes with GPU orchestration.
Education & Experience
- Bachelor’s, Master’s, or Ph.D. in Computer Science, Artificial Intelligence, Computer Engineering, or a related quantitative field.
- Demonstrated senior-level engineering experience (eligible for Level 20-IV or above) with extensive track record in high-performance software engineering and production AI deployment.