Develop Healthcare Workflow AI Copilot
Budget: ₹1,500 – ₹12,500 INR
I need an experienced AI/ML freelancer to take the lead on building a custom copilot that streamlines day-to-day clinical and administrative workflows for a healthcare platform. The immediate goal is to move from scattered manual steps to an intelligent assistant that can understand natural-language requests, pull data from our EHR, draft documentation, route approvals, and trigger follow-on actions inside existing tools.
You will have full freedom to choose the right blend of large language models, retrieval-augmented generation, and automation frameworks, provided the result is HIPAA-compliant, auditable, and easy for my in-house dev team to maintain. LangChain, FastAPI, vector stores such as Pinecone or FAISS, and orchestration services like Airflow or Prefect are all on the table if they help us reach production quickly.
To gauge fit, please share concrete past work that shows you have already delivered a similar copilot, assistant, or workflow automation project—ideally within a regulated environment. Live links, repo snippets, or short screen-captures are perfect; lengthy proposals are not required at this stage.
Acceptance criteria for the first milestone
• A running prototype accessed through a simple web interface or Slack/Teams bot
• At least three predefined healthcare workflows automated end-to-end (e.g., visit note drafting, prior-auth request, lab follow-up)
• Clear README with setup, environment variables, and model/embedding choices
• Security checklist confirming HIPAA-ready data handling and access controls
Once the copilot’s core loop is stable, we can expand into fine-tuning, evaluation, and full MLOps deployment, but the focus right now is that initial working assistant. I’m ready to start as soon as I find the right partner.
You will have full freedom to choose the right blend of large language models, retrieval-augmented generation, and automation frameworks, provided the result is HIPAA-compliant, auditable, and easy for my in-house dev team to maintain. LangChain, FastAPI, vector stores such as Pinecone or FAISS, and orchestration services like Airflow or Prefect are all on the table if they help us reach production quickly.
To gauge fit, please share concrete past work that shows you have already delivered a similar copilot, assistant, or workflow automation project—ideally within a regulated environment. Live links, repo snippets, or short screen-captures are perfect; lengthy proposals are not required at this stage.
Acceptance criteria for the first milestone
• A running prototype accessed through a simple web interface or Slack/Teams bot
• At least three predefined healthcare workflows automated end-to-end (e.g., visit note drafting, prior-auth request, lab follow-up)
• Clear README with setup, environment variables, and model/embedding choices
• Security checklist confirming HIPAA-ready data handling and access controls
Once the copilot’s core loop is stable, we can expand into fine-tuning, evaluation, and full MLOps deployment, but the focus right now is that initial working assistant. I’m ready to start as soon as I find the right partner.