AI Architect (Voice AI)
Neurons Lab · Remote · Poland · Portugal · Slovenia · Cyprus · Greece · Serbia · Hungary · Spain · Italy · Slovakia · Lithuania · Albania · Czech Republic · Moldova · Ukraine · Bulgaria · Estonia · Macedonia · Romania ·
COMPENSATION
€•• – ••k
Listed on the employer's job board (Ashby)
- Fully remote
- Voice AI & LLMs
- AWS & Python
- Senior level
- Starts Sep 2026
Join a project supporting a major US veterinary hospice network backed by private equity. You will own the technical architecture of a real-time voice copilot that assists care coordinators during live calls with pet families. The system extracts clinical data and updates scheduling systems instantly.
The role involves driving the solution into production, reducing latency, and ensuring strict data isolation. You will work closely with client stakeholders to maintain scope and deliver high-quality, evidence-based results while leading a small engineering pod.
What you'll do
- Own the full pipeline from streaming speech-to-text to AWS infrastructure
- Reduce P95 latency from ~6s to ~2s and resolve post-processing lags
- Conduct model A/B tests using golden-set evaluation for accuracy
- Monitor real per-call costs and optimize token economics via Langfuse
- Harden the system for 5–10+ concurrent calls with strict data isolation
- Lead the AI Engineer and manage tasks, reviews, and unblocking
- Present concrete system behavior and metrics to client stakeholders
What you'll need
- Mandatory experience shipping real-time voice or speech products to users
- 6+ years of hands-on AI/ML engineering with recent LLM production practice
- Proven record of reducing latency and fixing concurrency issues in live systems
- Strong Python skills and sufficient TypeScript for Chrome extension integration
- Experience with AWS Bedrock, serverless patterns, and SES
- Clear spoken and written English for communicating with US executives
- Ability to manage client expectations and handle detailed UAT scrutiny
Nice to have
- Experience with Chrome extension delivery mechanisms
- Familiarity with telephony stacks like Amazon Connect, Twilio, or LiveKit
- Production experience using Langfuse for observability
- Background in empathy-sensitive domains such as healthcare or insurance
- Experience working with US clients requiring Eastern time overlap
Stack
- Python
- TypeScript
- AWS
- Bedrock
- SES
- Langfuse
- Claude Haiku
- GPT Luna
How you'll work
- Fully remote position
- Start date expected in September 2026
- Initial commitment of 0.5 FTE, ramping to 1.0 FTE
- Multi-month duration with strong probability of extension
- Structured handover from outgoing architect included