Custom LIMS Development for SPEKTROKHIM: Automating Interlaboratory Proficiency Testing for 250+ Laboratories

Every quarter, one administrator at SPEKTROKHIM opened her inbox and started copying numbers.

250 laboratories. Each one emailed results in its own format. Some used commas as decimal separators. Some used dots. Some sent three files where one was expected. The administrator opened each one, found the right column, and pasted the value into the master spreadsheet.

One typo in that spreadsheet — one transposed digit — and the proficiency testing round loses credibility. Accreditation bodies don't accept "we fixed it later."

SPEKTROKHIM had been running this process on a legacy German LIMS platform for years. The platform was expensive, hard to support, and hadn't solved the core problem: results still came in by email, collation was still manual, and report preparation still took weeks.

They came to us with a clear brief: remove the manual layer entirely. Build something that's theirs. This case is how we approach custom LIMS development when a boxed product cannot cover the real workflow.

250+

active laboratories

600+

registered analysts

x3

faster reporting

x2

fewer corrections

How SPEKTROKHIM worked before

SPEKTROKHIM runs interlaboratory comparisons across Russia — petroleum products, food, water, industrial chemicals. The same sample goes to dozens of independent labs. Each lab measures it, submits results, gets evaluated against the cohort. That's how a laboratory proves measurement competence to accreditation bodies.

The German platform handled some of this. But "handled" is generous — it handled the parts it was designed for: internal lab workflows. Coordinating an external network of 250 independent participants was never what it was built for. That gap got filled with Excel and email, the same way every gap in every enterprise system gets filled.

By 2022, the workarounds had become the process. The platform was a liability, not an asset. And it was a foreign liability — licence renewals, vendor dependency, no control over the roadmap.

The brief was to replace it. Not adapt it. Replace it.

What we built

Organizer portal: running a round without the spreadsheet

The organizer portal is where SPEKTROKHIM's administrators live. We built it around one question: what does an administrator actually do during a proficiency testing round?

The answer: they set up the round, track who's submitted and who hasn't, chase the stragglers, close the round, and generate the report. Everything in between is coordination overhead.

The portal eliminates the overhead.

Rounds are created from a structured library of objects — crude oil, fuel oil, bitumen — and analytes: ash content, water mass fraction, sulphur content. Administrators pick from the library rather than re-entering methodology data every cycle. What used to take hours of setup takes minutes.

Participant management is live. The administrator sees, in real time, which labs have submitted, which are pending, which have flagged anomalies. No more emailing labs to ask if they're done. No more manually checking against a list.

When the round closes, the report exports to Excel with strict data types enforced — numbers stay numbers, decimal separators stay consistent. The audit trail is built into the export, not assembled from notes afterwards. Accreditation auditors get what they need without anyone preparing a separate document.

Lab portal: precision in, noise out

The participant-facing portal was designed around a single constraint: an analyst should be able to submit results without reading a manual.

Analysts log in and see exactly what they need to do — no more, no less. Their assigned programs, their open submission windows, their required analytes. Nothing else.

Result entry enforces precision at the field level. Values are accepted to five decimal places. An analyst cannot submit an incomplete form — every required field must be filled, every protocol scan must be attached. The system doesn't warn, it blocks. Only verified data enters the round.

This sounds strict. It is. That's the point.

Before this portal existed, the administrator's job included chasing labs for missing attachments, correcting format errors, and manually verifying completeness. The portal does that automatically. The administrator receives clean data or nothing at all.

Why custom, not off-the-shelf

This question comes up on every project. Here it had a clear answer.

Off-the-shelf LIMS products are built for internal laboratory workflows: a single lab, a single team, samples moving through a known pipeline. They're good at that. SPEKTROKHIM's problem isn't that — it's coordinating 250 independent external participants against a unified statistical framework. That's a different system: proficiency testing software and interlaboratory comparison orchestration, not a classic in-lab sample pipeline.

In regulated industries, a boxed LIMS also carries structural limits. Accreditation expects an auditable path from measurement to report. If your process spans dozens of external organisations, each with their own formats and habits, the product has to encode your rules — decimal precision, mandatory protocol scans, frozen data after round close — not a vendor's generic sample status model. Paying for configuration on someone else's architecture often means waiting on their roadmap while you still keep Excel as the real system of record.

The closest packaged products would have covered perhaps a third of the actual requirement. The rest would have been customisation: paying a vendor to build what you need, on their timeline, on their architecture, with their constraints. And then paying again when requirements change.

Custom lab software development costs more upfront. It costs less every year after. SPEKTROKHIM owns the system. No licence renewals. No foreign vendor dependency. When they need a new measurement method added, they add it — without filing a feature request. That independence is why custom LIMS development was the right call here, not another year of workarounds on a legacy platform.

Development process

Discovery. We mapped every step of a proficiency-testing round with SPEKTROKHIM: objects and analytes, participant onboarding, submission windows, corrections, export for audit. The output was a scope that reflected how rounds actually run — not a generic LIMS checklist.

Architecture. We designed the relational graph between objects, methods, national standards, and roles so organizers and labs could share one source of truth without leaking data across participants. Tables and exports had to stay correct under concurrent submissions.

Development. Backend, organizer UI, and lab portal moved in parallel with weekly demos. Edge cases from real rounds — incomplete files, mixed decimal formats, late submissions — were closed in the product, not left as "train the admin to fix it."

Rollout. Deploying a LIMS isn't just a code problem. It's a change management problem. Laboratory staff have muscle memory around their existing workflows — even bad ones. Training, phased go-live, and feedback from analysts on the bench mattered as much as the release build. We've written about why lab staff bypass new systems and how to prevent it in Why lab staff bypass LIMS. That work happened here too.

Stack and team

We needed a stack that stays stable under concurrent round submissions and doesn't buckle when administrators open reports with thousands of rows.

• Backend: Laravel — handles the complex relational graph between objects, analytes, methods, and national standards cleanly
• Frontend: Vue.js — responsive under heavy table operations, where administrators spend most of their time
• Database: MySQL
• Team: PM, frontend engineer, backend engineer, QA, designer

Result

More than 250 laboratories work through the platform today. Report preparation that used to take weeks of manual collation now finishes when the organizer closes the round — about three times faster.

The corrections rate dropped by half. Not because analysts got more careful — because the portal won't accept input that needs correcting.

SPEKTROKHIM owns an independent system built around their methodology. No legacy vendor. No Excel in the critical path. No administrator copying numbers from 250 email attachments.

Running a similar lab process?

If your team still moves proficiency testing, interlaboratory comparisons, or other regulated lab workflows through spreadsheets and email because nothing off the shelf quite fits — that's the gap we work in.

We develop custom LIMS and related lab platforms when the real process won't fit a boxed product. Start with how your rounds actually run; then engineer the system around that.

Discuss your project — more on WAPP lab solutions: chemical labs and LIMS.