Sustainability / CleanTech • ISO-Compliant GHG Reporting • Scope 1, 2 & 3
Confidentiality Notice: This case study discusses process decisions, architecture approach, and timelines only. No product features, user flows, or client-identifying details are disclosed. The client's intellectual property remains fully protected.
| | | |---|---| | Industry | Sustainability / CleanTech | | Compliance Framework | ISO-Compliant GHG Reporting | | Platform Type | Carbon Accounting & Energy Management | | Timeline | 2+ years (scope expanded with regulatory jurisdictions) | | Tech Stack | Node.js · React · TypeScript | | Key Integrations | IoT Energy Sensors · ETL Pipelines · AI Analytics | | Infrastructure | Self-Managed VPS · Docker · Database Replicas | | Recognition | Won Product of the Year in its category | | Key Result | 80% faster queries · 60% IoT reliability improvement |
The founder came to us with a clear vision: build a platform that gives businesses real-time, data-driven insights to reduce energy costs, track emissions, and simplify compliance reporting. They had evaluated off-the-shelf carbon accounting tools and found them costly, rigid, and fundamentally limited in how they handled data.
The core problem wasn't features, every carbon reporting tool has dashboards and charts. The problem was data architecture. Existing tools were built for Scope 1 and 2 emissions data, which is structured, predictable, and clean. The founder's ambition went further: Scope 3 supply chain data. And Scope 3 breaks everything.
Different suppliers report in different formats. Different time periods. Different levels of completeness. Some don't report at all and you need to estimate. Most sustainability platforms bolt Scope 3 on later. It never works because the underlying database schema can't handle the variability.
The founder needed someone who understood that the data model IS the product, and was willing to build the architecture for Scope 3 complexity from day one, not retrofit it later.
This project followed a principle that applies to every data-intensive build: design the data model for where the product needs to be in 12 months, not where it is today.
When the founder came to us, the initial approach was a standard NoSQL database, flexible, fast to prototype, and the default choice for most early-stage platforms. For Scope 1 and 2 data, NoSQL works fine. The data is structured. The queries are predictable.
But we looked at the actual requirements: ISO-compliant GHG reporting across Scope 1, 2, and 3. Real-time energy data from IoT sensors. Supply chain emissions from dozens of sources in dozens of formats. Analytics dashboards that needed to aggregate and slice this data in real time.
NoSQL would have worked for the first 6 months. Then the query performance would have collapsed under the weight of Scope 3 variability and the analytical workload.
The recommendation: start with NoSQL for ingestion flexibility, then build ETL pipelines to move processed data into a columnar database for analytics and reporting.
This wasn't a simple database swap. It was a fundamental architectural decision about how data flows through the platform, raw, messy supply chain data comes in through a flexible ingestion layer, gets normalised into a consistent internal format, then lands in a columnar store optimised for the exact query patterns that ISO-compliant reporting demands.
That architectural choice, made early, before the analytical requirements became urgent, is the reason the platform handles data that competitors choke on.
This is worth explaining in detail because it's the single biggest technical differentiator in sustainability tech.
Scope 1 emissions are direct, fuel burned, processes run. Clean data. Predictable format.
Scope 2 emissions are indirect from purchased energy. Still structured. Still manageable.
Scope 3 emissions are everything else in the supply chain. And this is where the data gets ugly.
A single company might have 50 suppliers across 12 countries. Each supplier reports emissions data differently, different formats, different time periods, different levels of granularity, different completeness. Some suppliers provide detailed breakdowns. Some provide a single annual number. Some provide nothing and you need to estimate using industry averages and allocation methods.
Most sustainability platforms handle Scope 1 and 2 beautifully, then fall apart at Scope 3. Why? Because they designed the data model for clean, structured emissions data. When messy, incomplete, multi-format supply chain data arrives, the schema can't handle the variability. The platform either rejects the data, requires manual cleanup for every import, or produces inaccurate reports.
We designed the data model for Scope 3 complexity from day one. A flexible ingestion layer that normalises messy supply chain data, regardless of format, frequency, or completeness, into a consistent internal representation. The schema expects the mess. It's architected for variability, not perfection.
This is why the platform won Product of the Year. Not because of features. Because the data architecture underneath handles real-world supply chain data that competitors can't.
The platform evolved through three distinct phases - PoC, MVP, and final production build - each adding architectural depth while maintaining the data model integrity established in phase one.
We migrated away from off-the-shelf products entirely. The existing tools were costing more in licensing and workarounds than a purpose-built platform would cost to develop. More critically, they imposed data structures that couldn't handle the Scope 3 requirements.
The custom build gave the founder complete control over customisation, security, and most importantly the data model. Every table, every relationship, every query path was designed for the specific analytical and compliance requirements of carbon accounting, not adapted from a generic SaaS template.
We selected, tested, and integrated energy IoT sensors for real-time consumption tracking and system performance monitoring. This wasn't plug-and-play, each sensor type produces data in different formats at different frequencies, and the platform needed to ingest all of it reliably.
The IoT integration improved system reliability by 60% and enabled real-time fault detection. When a sensor reports anomalous readings, the platform catches it immediately rather than discovering the data quality issue weeks later during reporting.
The key architectural decision: treating IoT data ingestion as the same pattern as Scope 3 supply chain ingestion. Variable formats, variable frequencies, variable reliability, normalised through the same flexible ingestion layer. One pattern, applied consistently.
This is where the early data model decisions paid off dramatically.
We implemented ETL pipelines to move processed data from NoSQL to the columnar database. The result: 80% reduction in query times for the analytical dashboards and ISO-compliant GHG reporting.
The ISO-compliant dashboards now produce accurate emissions tracking across all three scopes, in real time, not as a monthly batch process. Businesses using the platform can pull Scope 1, 2, and 3 reports at any time and know the data is current, accurate, and audit-ready.
If we'd built the analytics on top of the original NoSQL store, the typical approach, the platform would have hit a performance wall as data volume grew. The ETL pipeline architecture means the ingestion layer and analytics layer are decoupled. Each scales independently. Each is optimised for its specific workload.
The platform runs on self-managed VPS architecture - the same infrastructure philosophy as our healthcare builds, chosen because energy data and emissions reporting require full control over data handling.
The deployment architecture:
The infrastructure is future-ready, designed to support rapid growth in both data volume and user base without architectural rework.
The healthcare platform we built took 8 months. This one took 2+ years. Same budget class ($40K vs $50K). Same architectural discipline. Completely different timeline. Why?
Because compliance scope expanded with each regulatory jurisdiction added.
The initial build - PoC to production - moved fast. But carbon accounting regulations are not static. New jurisdictions adopt different reporting standards. ESG requirements evolve. The platform needed to support each new jurisdiction without rebuilding the foundation.
This is where the early architecture decision proved its value. The data model designed in month 1 is still the data model running in production today. No rebuilds. No rework. The foundation held through 2+ years and multiple jurisdiction expansions.
The longer timeline wasn't scope creep. It was scope expansion on a foundation built to handle it. That's the difference between a platform that grows and a platform that breaks.
80% faster queries for real-time analytics and reporting. The ETL pipeline architecture means ISO-compliant GHG dashboards return results in seconds, not minutes.
60% reliability improvement in IoT-based energy monitoring. Real-time fault detection catches data quality issues immediately rather than during monthly reporting cycles.
Product of the Year in its category. Not because of features, because of the data architecture underneath. The platform handles messy, incomplete, multi-format supply chain data that competitors reject or mishandle.
Significant cost savings from moving away from rigid off-the-shelf tools. The custom platform costs less to operate than the licensing fees for the tools it replaced, while handling data complexity those tools couldn't touch.
Future-ready infrastructure to support rapid data and user growth. The decoupled ingestion and analytics layers scale independently. New data sources, new jurisdictions, new reporting requirements, the architecture absorbs them without rework.
In sustainability tech, the invisible architecture is the competitive advantage. Users don't see the data model. They don't see the ingestion layer. They don't see the ETL pipelines.
They see that the platform works when others don't. They see that Scope 3 reports are accurate when competitors produce garbage. They see that a new supplier's data integrates in hours, not weeks.
That's not a feature. That's a data architecture decision made before the first line of feature code was written.
The most expensive decision in a sustainability platform happens in week one, the data model. If it's designed for Scope 1–2 simplicity, you'll rebuild when Scope 3 arrives. If it's designed for Scope 3 complexity from day one, the foundation holds for years.
This platform was designed for Scope 3 from day one. The foundation held. The product won an award. And the data model designed in month 1 is still the data model in production.
When I discuss past work, I share process decisions, architecture approach, and timelines. Never the product itself. Never the client.
I don't share what clients build. I don't name them. When I reference this project, I talk about how we structured the data architecture, what decisions mattered for Scope 3 complexity, and why the foundation survived 2+ years of expansion.
Your idea stays yours.
I run free 30-minute Build Plan sessions for founders who want a second opinion on their data architecture before they commit.
You share your product category and the data requirements you're dealing with. No product details needed. No IP shared.
You walk away with a 1-page decision doc: what's solid in your approach, what's risky, and a realistic timeline. Plus a build plan outline with phases. Two documents. Marked confidential. Yours to keep whether we work together or not.
DM me 'BUILD PLAN'
