Careers
Build the Future of Research Operations.
PetaSight is a software company building the Agentic AI OS for PMR execution. We are looking for people who want to build technology products that bring operational discipline to one of the most complex workflows in healthcare.
Open Positions
Full Stack AI Engineer
Mid / Junior
Hybrid – Bangalore, IndiaImmediate startCompetitive pay
We are looking for a Full Stack AI Engineer who is excited about taking research-backed AI systems to production. You will work closely with AI researchers to deploy and scale cutting-edge models and build reliable applications around them.
Requirements
- Demonstrable comfort using AI coding tools (Claude Code, Codex etc.)
- Proficiency in Python, Node.js, Next.js, JavaScript, SQL
- Ability to work independently and ship features end-to-end
What You Will Work On
- Deploying custom-trained models on AWS Bedrock
- Building production systems around LLMs
- Supporting research-to-product pipelines
- Collaborating with researchers to iterate quickly on product ideas
Responsibilities
- Develop and maintain frontend applications for AI-powered products
- Build and maintain scalable backend systems and APIs
- Find and use the latest, most appropriate AI tools and LLMs available
- Deploy and productionize custom AI models
- Deploy and manage cloud infrastructure (Bedrock models, GPU instances etc.)
- Implement reliable model serving, logging, monitoring, and eval pipelines
- Maintain clean, scalable, and well-documented code
Nice to Have
- Experience with Vercel, Lovable/bolt.new
- Experience with AWS cloud infrastructure (Bedrock, RDS, ECS etc.)
- Understanding of LLM training and fine-tuning
- Experience with self-hosted inference on NVIDIA GPUs
- Familiarity with model optimization and serving frameworks
- Experience with reinforcement learning for LLMs (DPO, PPO etc.)
- Infrastructure-as-Code experience (Terraform / OpenTofu)
- Experience working in startups or small engineering teams
What We Value
Governance First
Automation without accountability is risk. Every feature we build starts with the question: who approves, who is notified, and where is the evidence?
Operator Credibility
We build for the people who run studies, not the people who buy software. If it does not work in the field, it does not ship.
Ecosystem Respect
Research agencies, panel providers, and survey platforms are partners, not competitors. We orchestrate the ecosystem, not replace it.
Measured Outcomes
Every claim we make is tied to a measurable KPI. Cycle time, rework rate, defect escape rate, unit cost. If we cannot measure it, we do not promise it.
Interested in joining PetaSight?
Tell us what you would build and why you want to be part of this.