How to engage HL7 and FHIR developers in India
Decide first whether you need people or an outcome: a dedicated engineer working in your backlog, or a fixed-scope project per interface. Then test real integration judgement with a short trial on de-identified messages: an HL7 v2 feed with a Z-segment and a merge event, a FHIR search that pages, and a mapping specification to write. Before any patient data is reachable, settle named access through your SSO, no PHI on personal devices, and a BAA or data processing agreement covering the engineering work.
Healthcare integration is a narrow skill hiding behind common keywords
Search for HL7 and FHIR developers in India and you will find thousands of profiles listing both. Few of those engineers have been paged at night because an ADT feed stalled, untangled a patient merge that split a chart in two, or explained to a lab why its results were filed against the wrong encounter. Generic web developers can learn the FHIR REST API in a week. What takes years is the judgement around it: what a message means clinically, which fields a receiving system actually honours, and what to do when the specification and the live feed disagree.
This page is written for buyers in the US, UK, Australia and India who are weighing an Indian team for that work. It covers the roles, the ways to engage them, what public rate data does and does not tell you, the questions to ask any engineer and any vendor, the working-hour overlap you can realistically expect, and the access and contract terms that should be in place before anyone touches protected health information.
Radiatus is registered in Dover, Delaware and as an MSME in India. Our engineers work remotely from India; we have no premises for you to visit and do not pretend otherwise. Healthcare integration is our main practice, covering HL7 and EDI integration, FHIR interoperability and healthcare interfaces, and we place engineers through staff augmentation as well as running fixed-scope projects. Use the checklists below whichever vendor you choose.
The roles inside a healthcare integration team
"HL7 developer" covers at least five different jobs. Write your brief around the work, not the title.
| Role | Day-to-day work | Test them on |
|---|---|---|
| Interface analyst | Gathers requirements from both systems, writes the mapping specification, builds test messages, runs integration testing with the other side | Turning a vendor’s HL7 specification and three sample messages into a field-by-field mapping with open questions listed |
| HL7 v2 interface engineer | Builds and maintains channels in an integration engine (Mirth Connect, Open Integration Engine, Rhapsody, Iguana, Cloverleaf) or in code; MLLP, ACKs, transformations, error queues | Debugging a feed with escape sequences, repeating fields, a Z-segment and an out-of-order merge |
| FHIR API engineer | Builds FHIR clients and servers, SMART on FHIR apps and backend services, bulk data exports, validation against implementation guides | A paged search with _include, a conditional update with an ETag, and validating a resource against a profile |
| Integration architect | Chooses patterns, engines and hosting; owns identity matching, terminology, monitoring and the interface catalogue | Designing a results feed to three downstream systems with replay, monitoring and an outage plan |
| Integration DevOps / support | Runs engines and FHIR servers in production: upgrades, certificates, queues, alerting, on-call | An engine upgrade plan with rollback, and the alerts they would set on a stalled channel |
A small product team usually needs one FHIR API engineer and one HL7 v2 engineer who can also write mappings. A hospital programme adds an analyst and someone to run production. Ask any vendor which of these people you are actually getting.
Skills that matter more than the keyword list
- HL7 v2 depth. ADT events and what A08, A31, A40 and A44 do to the receiving chart; ORM/OML orders and ORU results; SIU scheduling; MLLP framing; original and enhanced acknowledgement modes; the escape sequences
\F\,\S\,\R\,\E\and\T\; how Z-segments break naive parsers. - FHIR in practice. R4 and where R4B and R5 differ; search parameters, paging and
_include; versioned references and concurrency withIf-Match; profiles such as US Core, UK Core, AU Base or the NRCeS guide for India; running the HL7 validator against an IG package in CI. - Security and identity. SMART on FHIR app launch and backend services (a signed JWT assertion exchanged for a token with system scopes), OAuth scopes, patient matching and why a demographic match is a clinical risk.
- Documents. C-CDA templates and sections, structured entries versus narrative, and converting C-CDA to FHIR without losing provenance.
- Terminology. LOINC for results, SNOMED CT for problems and findings, RxNorm or dm+d or AMT for medicines depending on the country, ICD-10 and its national variants, and mapping local codes with a reviewed table rather than guesswork.
- Country specifics. US: USCDI, information-blocking rules, X12 for claims and eligibility. UK: NHS number, MESH and NHS England’s FHIR APIs. Australia: My Health Record and AU Core. India: ABDM, ABHA and NHCX.
Engagement models for an Indian integration team
| Model | How it works | Best when | Risk to manage |
|---|---|---|---|
| Dedicated engineer (augmentation) | Named engineers work in your backlog, tools and stand-ups; you direct the work; billed by time | Ongoing interface work with changing priorities; you have an integration lead | Knowledge sits with individuals; insist on documentation and a replacement handover |
| Fixed-scope project | A defined interface or phase (for example an ORU results feed, or a FHIR API over your database) with acceptance tests and a fixed price | The scope can be written down; you want an outcome, not hours | Change control; agree how new message variants found in testing are handled |
| Pod (small team) | Analyst, engineers and a lead delivering a programme, such as an EHR migration or an HIE onboarding | Several interfaces at once with a deadline | Make sure your side still owns the interface catalogue and sign-offs |
| Support retainer | A monthly block of engineer time for fixes, partner onboarding and upgrades, with agreed response times | Interfaces are live and change slowly | Response times out of hours; what counts as a change rather than a fix |
| Direct hire or employer of record | You employ the engineer, directly or through an EOR | A permanent integration owner in India | Slow to find specialists; you carry continuity |
What does an HL7 or FHIR developer in India cost?
Public rate data for this specialism is thin, and much of what ranks for "rates" is written by vendors with a price to defend. What we could check on 13 November 2026:
- Arc’s page on hiring HL7 developers says HL7 developers on its marketplace "typically charge between $60-100+/hour (USD)", and that general freelance marketplaces list HL7 developers "as low as $10 per hour". Arc gives no India-specific figure.
- Freelance marketplace profiles and salary aggregators do show Indian HL7 and FHIR figures, but the samples we found were a handful of self-reported entries, so we do not repeat them as a benchmark.
The spread is real, and the cheap end is mostly generalists who list HL7 among twenty skills. Price is driven by three things: whether the engineer has run interfaces in production for a hospital or vendor, whether the vendor carries replacement and handover, and how much overlap with your hours you need. Radiatus does not publish a rate card. We estimate work in engineer-days, agree a fixed price per phase after a scoping call, and quote monthly terms for dedicated engineers once the role is clear.
| Typical task | Typical effort | What moves it |
|---|---|---|
| Inbound HL7 v2 ADT feed into an application, with mapping and tests | 5–10 engineer-days | Merge and update events; how many sending systems |
| Outbound ORU results feed to one partner | 8–15 engineer-days | Result types, attachments, the partner’s test cycle |
| SMART backend services client against an EHR’s FHIR API | 5–10 engineer-days | App registration and the EHR vendor’s review |
| FHIR API layer over an existing database | 20–40 engineer-days | Resources, profiles and search parameters required |
| Integration engine upgrade or migration, per ten channels | 5–15 engineer-days | Custom code in channels; test data available |
Questions to ask any HL7 or FHIR engineer
- An ADT^A08 arrives with a different date of birth for an existing patient. What does your interface do, and who should be told?
- How do you handle an A40 merge when the downstream system has no merge concept?
- A sender puts free text with a pipe character in OBX-5. What arrives, and how do you fix it at the source rather than in your code?
- When do you send an AE acknowledgement rather than an AR, and what happens to the message afterwards?
- Walk me through paging a FHIR search that returns 40,000 Observations. What breaks if you ignore the
nextlink? - How would you stop two systems overwriting each other’s updates to the same FHIR resource?
- Explain SMART backend services authorisation from key generation to the first API call.
- How do you validate a resource against a national profile in a build pipeline, and what do you do with warnings?
- A lab sends local test codes. How do you get to LOINC, and who signs off the mapping?
- Show me a mapping specification you wrote (de-identified). What was the hardest open question in it?
- What do you log from a message, and what must never appear in a log?
- Tell me about an interface you broke in production. How was it found, and what changed afterwards?
A four-hour trial task that works: give the candidate a de-identified ADT feed of 50 messages including one merge and one Z-segment, ask for a mapping to FHIR Patient and Encounter with tests, and a one-page note of questions they would put to the sending system’s team. The note tells you more than the code.
Questions to ask the vendor, not just the engineer
- Employment. Are the engineers your employees, or subcontracted? Will any work be subcontracted further without our written consent?
- Continuity. What happens if the engineer leaves: notice, overlap with a replacement, handover documents?
- Evidence. Can we see de-identified mapping specifications, test plans or runbooks from earlier work, and speak to someone who used them?
- Access. Will engineers work only through our SSO, VPN or virtual desktops, with named accounts and MFA? Which devices, and who manages them?
- Data. Where could PHI end up: laptops, chat tools, ticket attachments, AI coding assistants? What stops it?
- Contracts. Will you sign a BAA or data processing agreement covering the engineering work? Who owns the code and mapping documents?
- Entity. Which legal entity signs, in which country, and where are disputes heard?
- Claims. Ask for proof of every certification, partnership or client logo on the proposal. A specialist will have specific stories; a generalist will have logos.
Working-hour overlap: real windows for US, UK and Australian teams
Indian Standard Time is UTC+5:30 all year, so the overlap shifts twice a year with your clocks. The table assumes a 09:00–17:00 day on your side and the Indian shift shown.
| Your team | Indian shift (IST) | Shared hours, your winter | Shared hours, your summer |
|---|---|---|---|
| UK | 11:00–20:00 | 09:00–14:30 GMT (5.5 h) | 09:00–15:30 BST (6.5 h) |
| US Eastern | 14:00–23:00 | 09:00–12:30 EST (3.5 h) | 09:00–13:30 EDT (4.5 h) |
| US Central | 14:00–23:00 | 09:00–11:30 CST (2.5 h) | 09:00–12:30 CDT (3.5 h) |
| US Pacific | 16:00–01:00 | 09:00–11:30 PST (2.5 h) | 09:00–12:30 PDT (3.5 h) |
| Sydney, Melbourne | 07:30–16:30 | 12:00–17:00 AEST (5 h) | 13:00–17:00 AEDT (4 h) |
| Perth | 09:30–18:30 | 12:00–17:00 AWST all year (5 h) | |
Clocks change on different dates: the US on the second Sunday of March and the first Sunday of November, the UK on the last Sundays of March and October, and New South Wales and Victoria on the first Sundays of October and April. For a few weeks each year the gaps move by an hour and stand-ups need re-booking.
The gap has one real advantage for interface work. US and UK teams schedule cutovers, engine upgrades and partner go-lives out of hours to protect clinical systems; the US night falls inside the Indian working day, so an Indian team can run the change with full staff and hand over a written report when your morning starts.
PHI, BAAs and data protection
Most integration work can be done without patient data: synthetic messages, de-identified samples and test environments. Design for that first. When an engineer must see production data, for example to diagnose a failing feed, set the rules before day one:
- Access only through your environment (virtual desktop, bastion or VPN), with named accounts, MFA and logging; nothing copied to local machines.
- Production access time-boxed and approved per incident, with read-only rights by default.
- Logs, tickets and chat kept free of message bodies; screenshots of patient data banned.
- Engineers trained on the rules, with the training recorded.
For US work, a vendor whose engineers can access PHI on your behalf is a business associate (or a subcontractor of one) and should sign a business associate agreement. Where our engineers need PHI access, a BAA covering the engineering scope is agreed before work starts; we do not run your HIPAA compliance programme, risk analysis or audits, which is SemperWise’s work. In the UK, expect a data processing agreement under UK GDPR plus a transfer mechanism for India, such as the International Data Transfer Agreement. In Australia, Australian Privacy Principle 8 generally leaves you accountable for how an overseas recipient handles personal information, so the contract has to carry those obligations. In India, the DPDP Act and Rules apply; see our DPDP Rules checklist.
Red flags in a healthcare integration proposal
- "HL7 and FHIR" in a list of thirty technologies, with no named engine, message type or profile.
- No question back to you about sample messages, the other system’s specification or test environments.
- A fixed price for "integration with your EHR" before anyone has seen the EHR’s interface documentation.
- Partner badges or "certified" claims that the vendor cannot evidence.
- A plan to copy production data into the vendor’s own environment for testing.
- No mention of monitoring, error queues or what happens when a feed stops at night.
How Radiatus works with healthcare teams
Four ways to engage our integration engineers, all delivered remotely from India and scoped in engineer-days before any commitment.
Dedicated HL7 and FHIR engineers
The problem
Your integration backlog keeps growing and a permanent hire is months away.
How we do it
Named engineers from our healthcare practice work in your tools, backlog and stand-ups, directed by your lead, on the overlap window you agree. You interview them first; we handle replacement and handover if someone leaves.
What you get
- Engineers you interview and choose
- Agreed overlap window
- Documentation as part of the job
- Replacement with handover
Fixed-scope interface projects
The problem
You need a specific interface or API delivered, tested and handed over.
How we do it
A mapping specification first, then the build in your engine or code base, tests from de-identified samples, partner testing and a runbook. Priced per phase after the specification is agreed.
What you get
- Mapping specification
- Interface with automated tests
- Partner test support
- Runbook and monitoring
FHIR APIs, SMART apps and conversions
The problem
You need a FHIR API, a SMART on FHIR integration or data moved between HL7 v2, C-CDA and FHIR.
How we do it
FHIR servers and facades, SMART backend services and app launch, validation against national profiles in CI, and conversions such as HL7 to FHIR with reviewed terminology mapping.
What you get
- FHIR API or client
- Profile validation in CI
- Conversion maps with tests
- Developer documentation
Integration support retainer
The problem
Your interfaces are live, but nobody has time for partner onboarding, upgrades and the 2 a.m. stall.
How we do it
A monthly block of engineer time with agreed response times for fixes, new partners, engine and certificate upgrades, and monitoring of channel health.
What you get
- Agreed response times
- Upgrades and partner onboarding
- Channel monitoring and alerts
- Monthly report
Healthcare integration we build in our own products
Proof of the engineering, described exactly as far as it goes.
Radiatus OneHealth
Certified OpenEMR plus 25 practice modules, with nothing to migrate Our OpenEMR-based platform runs OpenEMR 8.2.0 unmodified and reads it through SMART on FHIR backend services (client_credentials with an RS384-signed JWT and system-level read scopes). It ingests HL7 v2 SIU scheduling messages (S12 to S26), and its engine-monitoring module watches Open Integration Engine, Mirth Connect up to 4.5.2 and BridgeLink, checking four alert metrics every two minutes. See the write-ups on our case studies page. Visit openemr.radiatus.comHow an engagement usually starts
Illustrative sequence for a dedicated-engineer engagement; fixed-scope projects follow the same first two steps.
Scope
Your systems, interfaces, partners and the overlap window you need.
Profiles and interviews
Engineer profiles matched to the roles above; you interview and can run the trial task.
Access and contracts
NDA, BAA or data processing agreement, SSO accounts, environment access, rules for PHI.
Onboarding
Interface catalogue, test environments, first small change merged and reviewed.
Review
Throughput, open risks, documentation added, and whether the team size still fits.
Standards, engines and tools our engineers use
What matters is which of these your project needs; ask any vendor to name the ones their engineers have used in production.
HL7 v2.x
ADT, ORM/OML, ORU, SIU, MDM and DFT over MLLP, with acknowledgements and error queues.
HL7 FHIR R4
REST, search, bulk data, SMART on FHIR, and national profiles including US Core and the NRCeS guide for ABDM.
C-CDA
Consolidated CDA documents and their conversion to FHIR.
Integration engines
Mirth Connect, Open Integration Engine and BridgeLink, with migration paths to and from other engines.
ABDM and NHCX
ABHA, health record linking and consent, and the claims exchange.
X12
837 claims, 270/271 eligibility and 835 remittance for US work.
Tools we use alongside it
- HAPI FHIR
- HL7 FHIR Validator
- Open Integration Engine
- Mirth Connect
- Postman
- PostgreSQL
- Docker
- GitHub Actions
Facts, versions and vendor details checked 13 November 2026.
Guides and related pages
Guides
Questions about HL7 and FHIR developers in India
Public data is thin. Arc says HL7 developers on its marketplace typically charge 60 to 100 US dollars an hour or more, and that general freelance marketplaces list some as low as 10 dollars an hour, but gives no India-specific figure. Price depends on production interface experience, the vendor’s replacement and handover terms, and the overlap you need. Radiatus estimates work in engineer-days and agrees a fixed price per phase after a scoping call.
An HL7 v2 developer builds message interfaces, usually in an integration engine, handling events such as admissions and results over MLLP with acknowledgements. A FHIR developer builds REST APIs and apps on FHIR resources, profiles and SMART on FHIR security. Many projects need both, because hospitals still run on HL7 v2 while newer systems expose FHIR.
With an Indian afternoon-to-late-evening shift, US Eastern teams typically share three and a half hours in winter and four and a half in summer; US Pacific teams get about two and a half to three and a half hours with a later Indian shift. UK teams can share five and a half to six and a half hours, and Sydney about four to five.
Yes, for the engineering scope we perform, where our engineers could access protected health information on your behalf. We prefer designs where engineers work with synthetic or de-identified data and reach production only through your environment. We do not run HIPAA compliance programmes, risk analyses or audits; that work belongs with SemperWise.
Use a short trial on de-identified data: about fifty ADT messages including a merge and a Z-segment, to be mapped to FHIR Patient and Encounter with tests, plus a one-page list of questions for the sending system’s team. Ask about acknowledgement modes, escape sequences, merges and how they keep patient data out of logs.
Radiatus is registered in Dover, Delaware and as an MSME in India. Our engineers work remotely from India, and we have no premises for clients to visit. Contracts can be signed with either entity depending on where you are.
Yes, as one of several integration engines alongside Open Integration Engine and BridgeLink, mostly for channel builds, upgrades, migrations and monitoring inside wider integration projects. If you are looking specifically for a long-term Mirth Connect support contract in the US, a dedicated Mirth support provider may suit you better.
Yes, with the right contract terms. UK work needs a data processing agreement under UK GDPR and a transfer mechanism such as the International Data Transfer Agreement. In Australia, Privacy Principle 8 generally keeps you accountable for overseas recipients, so obligations must be in the contract. Engineers must also learn the national profiles and services, such as UK Core or AU Core.
Need HL7 or FHIR engineers who have run interfaces in production?
Tell us your systems, the interfaces or APIs you need, your time zone and whether you want dedicated engineers or a fixed-scope project. A senior engineer will reply with the roles we would propose and a typical effort range in engineer-days. Or book a free 30-minute call.






Compliance-first delivery