What You’ll Build & Own
Own the design and delivery of production-ready AI features for healthcare use cases across clinical, home care, operational, and patient-facing workflows in the US market.
Build and improve agentic AI systems, including multi-agent orchestration, autonomous tool-use workflows, and medical reasoning pipelines.
Develop and maintain high-performing RAG systems using clinical data, medical records, and complex healthcare data structures.
Lead model adaptation and fine-tuning efforts using methods such as LoRA and QLoRA on open-source models like Llama and Mistral.
Improve AI workflow performance across latency, reliability, token usage, cost efficiency, automated evaluation quality, memory management and context management
Write clean, scalable, secure backend code with a strong focus on async systems, data integrity, compliance, and real-world clinical usage.
Work closely with founders, product teams, healthcare experts, and US-based stakeholders to convert ambiguous healthcare ideas into scalable AI products.
What We’re Looking For




6–9 years of software engineering experience, especially in backend systems, async architectures, and complex data pipelines.
Strong experience shipping AI products to real users, with attention to edge cases, deterministic testing, performance, and scale.
Deep understanding of agentic AI concepts, including tool calling, state management, multi-agent coordination, and self-correction loops.
Hands-on experience with PEFT, dataset preparation, and fine-tuning open-source models for domain-specific tasks.
Familiarity with US healthcare workflows, sensitive data handling, and structured clinical data formats, or a robust interest in learning them deeply.
Expert-level Python skills, strong asyncio experience, robust API development experience, and exposure to cloud platforms such as AWS, GCP, or Azure.
Startup-oriented mindset with strong ownership, speed, practical decision-making, and the ability to drive outcomes ind

📌 AI Engineer (Chennai)
🏢 OSpectra AI
📍 Chennai