Data Engineer – Healthcare Data Platform (Provider Analytics)

Neolytix; Gurugram District; On-site; INR 2,500,000 - 4,000,000; Full time

Job summary

Neolytix is building a healthcare data platform to serve client reporting and the InCredibly credentialing solution. The role centers on delivering lakehouse architectures, ingestion pipelines, and a robust validation layer, with growth toward independent platform ownership within 6–12 months.

You will own the data pipelines from ingestion to production and work closely with the Lead Architect to implement automated validation, governance, and scalable data models across the RCM domain.

Qualifications

• Proficient in SQL and Python for building data pipelines.
• Experience with cloud data platforms and lakehouse concepts.
• Knowledge of data governance and validation practices.
• Familiarity with HIPAA/PHI handling and security practices.
• Ability to own end-to-end data products from ingestion to production.

Responsibilities

• Own pipeline end-to-end: ingestion, transformation, validation, and production readiness.
• Document architecture and create runbooks to enable future ownership.
• Integrate new data sources via API or structured feeds as clients onboard.
• Implement data governance: lineage, validation, and access controls.
• Collaborate with analytics and product teams to deliver trusted data for reporting.

Skills

Strong SQL

Python for data engineering

Analytical mindset

Problem solving

HIPAA/PHI data handling

Tools

Azure Data Factory

Azure Synapse

Databricks

Power BI

ETL pipelines

Job description

Data Engineer — Healthcare Data Platform

Reports to: Lead Architect, Provider Analytics

Location: Gurgaon (Office-based)

Experience: 4–7 Years

Level: Mid‑Senior IC

Neolytix is standing up a dedicated analytics function that owns all client‑facing reporting for our RCM business (20 clients today) and the data foundation behind InCredibly, our credentialing platform. Reporting is moving out of Operations and onto an engineered data platform with formal governance: automated extraction pipelines, a certified metric layer, and validation built into the data flow rather than bolted on afterward. Platform architecture is owned in‑house by a hands‑on data architect who sets the design — medallion lakehouse structure, modeling standards, and the validation framework. This role builds the platform inside that design, owning pipelines end‑to‑end from ingestion through production operation, and is explicitly expected to grow into independent ownership of the platform within 6–12 months.

Mission

Build Neolytix’s healthcare data platform inside an architect‑defined design: the lakehouse, ingestion pipelines, data models, and automated validation layer that power client reporting across our RCM book and the data needs of the InCredibly platform. Within 12 months, the majority of recurring client reporting runs on pipelines you built and operate — with data quality proven by reconciliation, not assumed — and you are running the platform with review‑only oversight.

What You Will Build (First 12 Months)

• The medallion lakehouse (bronze / silver / gold) on the Microsoft data stack, implemented to the architect’s design and documented so every decision is reproducible
• Automated extraction connectors for EHR / practice‑management systems (target: 4 systems in year one), plus ingestion of clearinghouse 835 / 837 files and client‑supplied feeds — replacing manual monthly pulls
• A claim lineage database linking charges, claims, remittances, and payments across sources (target: ≥90% linkage rate), enabling true first‑pass resolution and denial analytics
• Dimensional (star‑schema) models for core RCM KPIs — first‑pass resolution, days in AR, denial rates, net collection rate — serving certified Power BI semantic models
• The automated validation harness inside the pipelines: row counts, reconciliation‑to‑source totals, period‑over‑period variance thresholds, and alerting — so errors are caught before a client sees them
• A bounded InCredibly workstream: support for an in‑flight client data migration and integration groundwork (API / FHIR) for the platform’s reporting needs

Ongoing Responsibilities

• Own pipeline reliability end‑to‑end: orchestration, monitoring, failure recovery, schema‑drift handling, and refresh SLAs
• Absorb and document the architecture as it is built — decision records, runbooks, and standards — with the explicit goal of independent platform ownership within 6–12 months
• Implement the technical half of data governance: lineage tracking, enforcement of certified metric definitions in the transformation layer, and role‑based access to data assets
• Integrate new data sources as clients onboard — API‑based where available, structured file exchange (SFTP) where not
• Handle PHI to HIPAA standards: minimum‑necessary access, encryption in transit and at rest, and audit‑ready handling in every pipeline
• Work day‑to‑day with the associate data engineer; grow into a mentoring role as ownership expands

What You Bring

Must‑Have

• Strong SQL and solid Python for data engineering
• Hands‑on Azure data stack experience — Azure Data Factory plus Synapse, Fabric, or Databricks (equivalent lakehouse experience on another cloud considered with strong fundamentals)
• Has owned at least one pipeline or data product end‑to‑end — contributed to its design, built it, and operated it in production — as opposed to executing assigned tickets inside someone else’s build
• Working knowledge of dimensional modeling applied in production
• A demonstrated validation and reconciliation mindset: you can explain, concretely, how you knew your numbers were right. This is non‑negotiable at any level
• Disciplined handling of sensitive / regulated data; HIPAA or PHI exposure preferred
• Evidence of fast learning and increasing scope — this role is a succession seat, and trajectory matters as much as current state

Strong Pluses

• US healthcare claims and remittance data (835 / 837), EHR / practice‑management exports, or RCM metrics exposure
• FHIR / HL7 integration experience; REST API development
• Power BI semantic‑model / dataset design (the front end sits on your models)

What Success Looks Like

• 90 Days: First client pipeline live end‑to‑end under architect review; validation harness v1 implemented and catching reconciliation breaks automatically.
• 6 Months: Connectors and the claim lineage database in production; operating pipelines with review‑only oversight; manual extraction hours measurably reduced.
• 12 Months: 4 EMR/PM connectors live with a repeatable playbook; ≥90% claim linkage; ≥97% scored data accuracy on certified reports; independently owning platform run and extension.