ML-AI Engineer for an innovative med-tech start-up

Náplň práce, právomoci a zodpovednosti:

Seerlinq is a medical telemonitoring platform. This role owns the AI layer across the product.Three active work areas:

1.  Medical document processing — multilingual medical data extraction from PDF / photo records

2.  Voice module — multilingual voice interface to report vitals and symptoms

3.  Clinical AI agent — an LLM-powered assistant for clinicians

Responsibilities:

• Design and deploy the medical document processing service: DB migration, dedup logic, harden the doctor-in-loop review flow

• Implement the voice pipeline (Phase 1: SaaS STT/LLM/TTS providers, WebSocket + HTTP transport, server-side VAD, GDPR consent layer)

• Maintain and extend the clinical AI agent and its tool integrations

• LLM integration patterns: structured JSON output, streaming, provider abstraction (SaaS ↔ self-hosted OSS)

• Contribute to multi-region deployment readiness where it touches AI services

• Write tests, maintain async background workers, write database migrations

Why join Seerlinq?

• Mission-driven, experienced collegues with strong academic and technical backgrounds

• Secured grant to expand globally

• High impact in early stage startup, greenfield AI projects

Zamestnanecké výhody, benefity:

• A dynamic role in a future-oriented company

• A young and friendly team

• A modern and comfortable work environment

• Continuous opportunities for professional and personal growth

• Care for your health 

• Company events and team-building activities

Informácie o výberovom konaní:
If you're interested, feel free to contact us by responding to the advertisement or via email at: careers@seerlinq.com

Požiadavky na zamestnanca:

Current / Proposed Stack:

Python 3.13+, FastAPI, SQLAlchemy (async), PostgreSQL + pgvector, llamaindex, Redis, Taskiq, Ollama, SGLang/llama.cpp, pydantic-ai, BGE-M3 embeddings, Docker, Kubernetes (EKS), Helm, Hugging face

Required

• Python async programming (FastAPI, asyncio)

• LLM API integration — structured output, streaming, prompt engineering

• Vector databases / RAG pipelines

• Docker + basic Kubernetes operations

• Proven track record

Nice to Have

• Speech processing: STT/TTS pipelines, WebSocket audio streaming, server-side VAD

• NLP hands-on (embeddings, seq2seq, GPT, tokenization, transformers)

• MLOps hands-on (training pipelines, deployment efficiency, monitoring, observability)

• pydantic-ai or comparable agentic framework

• AWS (EKS, SQS, S3, Secrets Manager)

• Medical/clinical data handling (GDPR, PHI awareness)