Voice AI Infrastructure & Integration Engineer
India · À temps plein
Soyez le premier à postuler
- Expérience
- 5–8 yrs
- Salaire
- INR 3,000,000 – INR 3,500,000 / year
- Ouvertures
- 1
- Publié
- il y a 3 heures
- Work mode
- Au bureau
- Éducation
- Bachelor’s or Master’s in Computer Science, Information Systems, or related field
- Eligibility
- Candidates with 5 to 8 years of relevant experience who can join immediately and have practical experience in Voice Bots or Voice AI products may apply. Backgrounds in DevOps, SRE, platform engineering, or real-time telephony systems are especially relevant. Applicants should be qualified with a Ba…
- Resume
- Required to apply
Where you'll work
Description de l'emploi
Role Summary
This position is for a Voice AI Infrastructure & Integration Engineer with 5 to 8 years of experience. The role focuses on the reliability, monitoring, scaling, and telephony connectivity of an LLM-based voice agent platform. The selected professional will also manage call recording workflows, compliance requirements, orchestration across containerized systems, and integrations with telephony and post-call business systems.
Compensation for this opening is budgeted at ₹30 to ₹35 LPA, depending on how closely the candidate’s background matches the role. Only immediate joiners are being considered.
Mandatory Requirement
Applicants must have practical, hands-on experience building Voice Bots or Voice AI products and solutions.
Observability, Logging & Monitoring
- Build and maintain an end-to-end observability stack using tools such as OpenTelemetry, Datadog, and the ELK stack so traces, logs, and metrics are visible across the system.
- Instrument the voice workflow to measure per-turn performance indicators such as STT latency, LLM inference duration, guardrail evaluation time, TTS first-chunk latency, and complete round-trip time.
- Create alerts for SLA breaches, spikes in guardrail triggers, increases in STT/TTS failures, and telephony drop rates.
- Develop live dashboards for call traffic, simultaneous sessions, and vendor health.
Call Recording & Compliance
- Own the recording workflow for calls, including full-duplex audio capture and turn-level transcripts, with storage in S3 or Azure Blob under FDCPA-compliant retention rules.
- Connect call data to QA and audit tools such as MaestroQA or EvaluAgent for playback, AI-supported scoring, and compliance review.
- Protect stored audio under PCI-DSS requirements by masking payment card content before long-term retention.
CI/CD, Deployment & Scaling
- Create CI/CD pipelines for orchestrator settings, system prompts, Markdown scripts, guardrail rules, and infrastructure-as-code assets.
- Package voice pipeline components in Docker and deploy them on Kubernetes platforms such as EKS, AKS, or GKE with autoscaling based on call volume.
- Use blue-green and canary release approaches to achieve safe, zero-downtime deployments.
- Maintain session-state infrastructure using Redis or Upstash with expiry, TTL, and failover support for per-call state.
Telephony & Post-Call Integration
- Integrate with telephony providers like Twilio, Vonage, and Plivo for SIP trunking, call routing, and PCM or μ-law audio handoff.
- Configure call windows, time-zone logic, contact-frequency limits, and Do Not Call enforcement at the telephony layer.
- Build asynchronous post-call workflows such as CRM updates, email confirmations, call summaries, and payment gateway interactions.
- Implement warm transfer flows so calls can be escalated to human agents with complete conversation context.
Experience and Background
The role calls for 5 to 8 years of experience overall, including at least 3 years supporting real-time or voice/telephony systems. Candidates should have a history of working on production observability systems for distributed, low-latency environments. Strong Kubernetes knowledge is expected, including autoscaling, GPU node management, and multi-service deployments. Familiarity with SIP, RTP, WebRTC, and at least one CPaaS provider is required.
Preferred Background
Experience in BPO operations, contact centers, or financial services voice AI infrastructure is preferred. Exposure to GPU infrastructure such as NVIDIA A100/H100 or Triton Inference Server, as well as orchestration platforms like Pipecat, LiveKit, or Vocode, will be an added advantage. Experience with cost optimization for live AI voice systems is also useful.
Education
A Bachelor’s degree or Master’s degree in Computer Science, Information Systems, or a related discipline is preferred.
Submission Instructions
- Candidate profiles must be uploaded only through SuccessFactors against the relevant Job ID.
- Resumes should not be sent directly by email.
- After uploading a profile in SuccessFactors, candidate details should be shared in the required tracker by email.
Contact
For coordination, connect with Abhishek at abhishek.m@livecjobs.com or via WhatsApp at 9154908075.