Senior Software Engineer, LLM & Auth Letter Pipeline
Build the AI pipeline behind eligibility and authorization: real systems, real PHI boundaries, real production traffic.
The role
What this job actually is
You build and own services in our eligibility and authorization AI stack: parsing payer responses, generating authorization documentation, and automating what used to be phone calls. This is production software handling healthcare data, so the PHI boundary is a design constraint, not an afterthought. You will work with people who understand the billing domain deeply, which makes the requirements unusually concrete.
Ownership
What you will own
Pipeline services
Build and maintain the services that parse, transform and act on payer data.
LLM integration
Design prompts, guardrails and evaluation so model output is reliable enough to act on.
PHI-safe design
Keep protected health information inside the boundary by design. This is non-negotiable.
Evaluation
Build the harness that proves a change improved accuracy rather than assuming it.
Reliability
Own uptime, error handling and observability for what you ship.
Collaboration with operators
Work directly with billing specialists to encode what they actually know.
Ramp
Your first 90 days
First 30 days
Learn the stack, the domain and the PHI boundary. Ship a small change to production.
By day 60
Owning a service, with evaluation in place for the changes you make.
By day 90
Delivering meaningful automation improvements measured against a baseline.
Requirements
What we look for
- Five or more years building production software, strong Python
- Experience integrating LLMs into production systems including prompt design and evaluation
- Solid data pipeline and API engineering skills
- Understands why handling sensitive data carefully matters
- Able to work directly with non-engineers to capture domain rules
Bonus
Strong plus, not required
- Healthcare data experience including X12 270/271
- Experience with document parsing and OCR
- Background in evaluation harnesses for model output
Accountability
How this role is measured
| Measure | What it means |
|---|---|
| Automation rate | Work handled without human touch. |
| Output accuracy | Measured against a reviewed baseline. |
| Reliability | Uptime and error rate. |
| Cycle time | Idea to production. |
| PHI boundary integrity | Zero violations. |
| Operator satisfaction | Does it actually help. |
Why ASP-RCM
Why this seat is different
We are senior-led by design. A senior partner is named on accounts and shows up in the work, which is why clients renew and why this role carries real authority rather than a title. Our AI suite is in production, not in a roadmap deck, and our compliance posture is independently audited: SOC 2 Type II with HITRUST, ISO 27001 and HIPAA.
We pay for certifications and run upskilling cohorts, because coding and revenue cycle are crafts that need investment. You will work across Dallas HQ, six US states and Chennai, India, with colleagues who have run these functions at scale.