AI Prompt Engineer (Internal AI Tooling & Automation)

Level

Individual contributor (mid to senior)

Location

Noida, India (on-site)

Employment Type

Full-time, XpertDox India Private Limited

Compensation

INR 30-50 LPA

About XpertDox

XpertDox builds XpertCoding, an autonomous medical coding platform. XpertCoding codes claims end-to-end using a proprietary AI engine built in-house, running at above 94 percent automation and above 99 percent code-level accuracy in production. We operate across Scottsdale, Arizona and Noida, India, in a regulated healthcare environment under HIPAA, SOC 2 Type 2, and ISO 27001.

Position Summary

This role builds the shared layer of AI-powered automation and tooling that makes engineering across XpertDox faster and more consistent.

You will spend most of your time building: reusable AI projects and skills for recurring technical workflows, integrations that connect internal systems to AI tooling, agent configurations tuned to our stack, and automation for internal processes that currently consume engineering time. You document and hand off what you build so others can adopt it without you in the loop.

This is a hands-on individual-contributor role for a solid, self-directed builder. It is not a management or strategy role.

Responsibilities

Build AI projects and skills

- Build and maintain reusable AI projects and skills for recurring technical workflows: document generation, rule authoring, code-review support, mockup production from specifications, and analytics reporting.
- Follow clear architectural conventions: one responsibility per skill, shared logic factored out, explicit inputs and outputs so pieces compose predictably.
- Keep project instructions and skill definitions in source control, under the same review discipline as application code.




- Track which artifacts are actually used, refactor the ones that have grown unwieldy, and retire the ones that have not earned their keep.

Build tooling integrations

- Build and maintain MCP servers that connect internal systems (source control, ticketing, business intelligence, and internal documentation) to AI tooling.
- Maintain house-standard configurations for agentic AI coding tools, including project-level instruction files and repository conventions.
- Report usage and spend across AI tooling so priorities and budget can be set with real numbers.

Build internal automation

- Build agentic workflows to address internal bottlenecks identified with leadership and function leads.
- Instrument what you build so its effect is measurable rather than assumed.
- Build evaluation suites for the tooling you deliver: define test cases, scoring criteria, and pass thresholds, and version them alongside the configuration they test.

Document and hand off

- Write and maintain practical reference material: worked examples and pattern documentation drawn from our own codebase, so what you build can be adopted independently.
- Provide written guides that work asynchronously across the India and US time zones.

Boundaries You will not own the production AI and ML coding models; that work sits with a separate function. This role governs internal development tooling and automation, not product model architecture.



You will not set organizational AI governance policy. Firm guardrails apply to PHI exposure paths across all AI-assisted workflows, and they are set for you rather than by you.

Qualifications

Required

- Substantial working use of agentic AI coding tools such as Claude Code or Cursor. You should be able to describe specific things you have built with them, including where they broke and what you changed.
- Practical experience designing prompts and context for reliability, with a working understanding of failure modes rather than a collection of techniques.
- Strong Python and comfort with JavaScript or TypeScript.
- Strong self-direction: able to identify high-value automation opportunities and ship them with minimal oversight.
- Clear written communication, since a meaningful share of your output is documentation read asynchronously across two time zones.

Preferred

- Experience building MCP servers or similar tool-integration layers for LLM systems.
- Experience building internal tooling used by other engineers, at any scale.
- Exposure to evaluation design: measuring whether an AI system works rather than judging by impression.
- Experience in a regulated environment, particularly healthcare or HIPAA.
- Experience working across US and India time zones.

Not a Fit

- Candidates whose AI experience is limited to using tools as a consumer, without building systems around them.
- Candidates looking for a management or team-lead track in the near term, since this role is deliberately hands-on and individual.
- Candidates seeking to work on production model architecture, which sits with a different function.

Ready to join our team? click below to apply:

https://careers.xpertdox.com/

📌 AI Prompt Engineer (Noida)
🏢 XpertDox
📍 Noida