Business Problem
Solar operators were dependent on bundled hardware-and-software ecosystems. A vendor-agnostic dashboard would reduce lock-in and centralize plant visibility.
What this was
Not a product launch. A serious proof.
The solar monitoring work was intended as a commercial opportunity, not a toy project. Budget and timeline conversations happened, and the end goal was a deployable vendor-agnostic dashboard. The project did not convert into a contracted commercial rollout, so it should be presented honestly as a commercial-intent proof of concept.
Its value is still real. It was the first project that forced Salil into industrial systems: hardware, site conditions, sparse device documentation, cloud ingestion, payload parsing, database structure, dashboards, and the gap between what vendors say a device exports and what it actually sends.
How it began
Field first, dashboard later
Internship
Warm introduction
Sanjay connected Salil to Prashant Pai, CEO/MD of G S Solar Systems. The first work was practical, not theoretical: hardware installation, DG meters, IoT sensors, and connectivity testing.
Rohtak
Client-site installation
Salil visited K Automatic factory in Rohtak, installed DG meters, and tested whether the deployed hardware could communicate reliably from a real industrial site.
Sriharikota
Live plant study
A five-day trip to Sriharikota exposed Salil to a full solar plant environment. Daily reports were prepared under Prashant Pai's supervision while studying how generation, inverter data, and plant operations worked in practice.
Coffee discussion
The actual problem emerged
Prashant showed the existing monitoring product and the dependency problem: the software was tied to the hardware provider. The desired platform was a centralized, vendor-agnostic login and dashboard for monitoring live power generation.
Discovery
Wattmon payload discovery
The primary obstacle was the data logger. Documentation was limited and outdated. The expected CSV stream turned out not to be the real integration format; after weeks of log reading and parsing, the actual payload was found to be form-urlencoded request data.
POC
Cloud pipeline built
The working proof used AWS API Gateway, Lambda, a payload transformation layer, MongoDB cloud storage, and a crude dashboard. It was not elegant, but it proved that live plant data could be captured and visualized outside the vendor's locked stack.
This project taught systems architecture the hard way: hardware, bad docs, payloads, cloud functions, databases, and live industrial telemetry.
Technical Problem
The data logger integration was poorly documented. The assumed CSV export did not match the live request format, forcing manual discovery and payload parsing.
Architecture Lesson
The first working version used what was available and understandable at the time: API Gateway, Lambda, transformation templates, MongoDB, and a rough dashboard.
What was proven
POC outcome
- Live solar telemetry could be routed outside the vendor-provided SaaS stack.
- Wattmon logger payloads could be interpreted despite weak public documentation.
- A cloud ingestion path could receive and store plant readings.
- MongoDB could be used as a fast learning path for structured telemetry storage.
- A rough dashboard could visualize live plant data from the Sriharikota factory context.
What not to overclaim
Honest status
- This was not a contracted commercial deployment.
- It was not a mature SaaS product.
- It was not architected with the polish Salil would use today.
- It was a commercial-intent proof of concept that validated feasibility.
- Its real value was the field learning and architecture confidence it created.
Outcome
Why this case matters
The Solar Monitoring POC belongs in the portfolio because it explains where Codeforge's industrial-software instincts came from. It shows hands-on exposure to hardware, factories, field installation, solar plant operations, cloud ingestion, and messy real-world data. It is not the biggest commercial win. It is the origin story for later confidence in IoT dashboards, enterprise systems, and operations software.