Forem: Hargun Kaur

🚀 My Journey into the Engineering for Change Fellowship: Engineering for People and Planet

Hargun Kaur — Tue, 09 Dec 2025 22:48:21 +0000

Engineering has always excited me, but what truly motivates me is building solutions that make life better for real people. This motivation led me to one of the most meaningful milestones in my career: being selected for the Engineering for Change Fellowship. A global program where engineering talent meets social innovation.

This fellowship provided me the chance to work on technology that uplifts communities, especially those often overlooked. And the journey to get here was as impactful as the work itself.

🌍 What is Engineering for Change?

Engineering for Change (E4C) is a global initiative powered by nonprofits and foundations including ASME, Siemens Foundation and Autodesk Foundation. The fellowship connects engineers, designers and researchers from around the world to build sustainable, inclusive and locally relevant solutions.

Their north star is clear. Engineering should serve all of humanity. Not just the privileged few.

🎯 My Application Experience

The selection process is competitive since applicants come from over 100 countries. But what makes it truly different is that it looks for your purpose, not just your resume.

• I shared my passion for accessible and impactful technology and selected the organisations that I was interested in.
• I participated in an interview focused on values like empathy, community listening and sustainability

The day I received my acceptance mail was unforgettable. It felt like a validation that I was on the right path. A path where technology and humanity walk together.

💰 The Fellowship Stipend

Yes, this is a paid global fellowship. E4C ensures fellows are supported while contributing to real-world problem solving. The stipend averages around 3000 USD for the full term over a period of around 5 months but varies slightly depending on your geographical location and cost of living.

It is not a traditional job paycheck. It is a recognition that meaningful innovation deserves support.

🌱 My Project: Empowering Farmers Through Technology

During my fellowship, I contributed to a project focused on strengthening agricultural ecosystems and improving farmers access to essential services.

Farmers often struggle with:

• Fair pricing for produce
• Access to inputs and advisory services
• Awareness of digital tools available to them
• Climate related uncertainties and resource limitations

Our goal was to help design a framework for digital agriculture solutions connected to open public infrastructure. This included analyzing the needs of farming communities, mapping services and proposing systems that are:

• Easy to adopt
• Affordable for smallholder farmers
• Scalable and interoperable
• Built for local realities, not external assumptions

It showed me how deeply technology must integrate with cultural, financial and environmental contexts to truly work on the ground.

🎓 Learning Every Week with Global Experts

One of the most exciting parts of the fellowship is the learning journey. Every week, we engage in curated modules that explore real challenges: climate resilience, circular economy, sustainable agriculture, affordable housing, and more.

We also attend live talks from speakers who are not just experts but changemakers shaping sustainability in the real world. People leading social enterprises, engineers building climate tech, researchers improving access to health and clean water. Their perspectives show how innovation can reduce inequity and transform systems.

And the best part: you network with brilliant, like-minded fellows from every corner of the globe. People who care about purpose. People who stay awake at night thinking about impact. You learn from them as much as from the curriculum.

🤝 What Makes E4C Special

It is the way E4C treats communities as problem solvers, not just beneficiaries. Instead of building for people, we build with them.

During the fellowship you learn about:

• Human-centered design
• Sustainability and lifecycle thinking
• Partnerships with purpose
• Equity and ethical responsibility in engineering

You do not just design. You engage. You listen. You learn.

🔮 The Road Ahead

The fellowship might have ended but the mission continues. The experience shifts how you see engineering. Every project now comes with questions like:

Does it empower people?
Is it sustainable for the planet?
Will it remain useful long after deployment?

There is a new standard for what it means to call something a solution.

🌟 Final Message

If you are someone who feels engineering can be a force for good. If you want your skills to help real communities thrive. If you want to work on global problems with real-world partners.

Keep an eye on E4C. The next fellow could be you. Reach out to me on Linkedin if you have any questions, and I’d be more than happy to help you out! :)

Engineering becomes meaningful when it touches lives. I am grateful that this fellowship allowed me to step into that purpose.

Automating Healthcare Workflows: Onco-Assist

Hargun Kaur — Mon, 01 Sep 2025 06:54:54 +0000

This is a submission for the AI Agents Challenge powered by n8n and Bright Data

What I Built

When technology meets personal need, innovation becomes deeply meaningful. This project wasn’t just a hackathon build or an experiment in automation — it started from my own mother’s case. Managing her medical reports, prescriptions, test results, and hospital bills was overwhelming. Each piece of information lived in different silos — Gmail attachments, WhatsApp messages, portals — leaving us scrambling whenever a doctor asked for history. I wanted to solve this once and for all.

That’s when I turned to n8n, an open-source workflow automation tool, and started building an automated pipeline to collect, classify, process, and deliver her medical information in an organized way.

Demo

https://drive.google.com/file/d/1e5ZRD-Kij_X26GvbFbSV6DXEGblp9JmU/view?usp=sharing

n8n Workflow

https://github.com/hargunkaur286/onco-assist

Technical Implementation

📨 Step 1: Gmail Integration — The Entry Point

Every journey starts with an email. Most of my mother’s medical information — PET/CT reports, blood tests, invoices, prescriptions — came into Gmail as attachments. I set up a Gmail Trigger node to fetch every new email with a medical-related attachment. This ensured nothing slipped through.

Why? Because manually downloading and renaming files was error-prone. Automating this saved us hours every week.

🧠 Step 2: Text Classifier — Sorting the Chaos

Once emails were captured, the next challenge was classification. A PET scan report looks very different from a pharmacy invoice, and handling them required different workflows.

I used n8n’s AI Text Classifier with an LLM backend. It looked at the subject line and email body to classify attachments into buckets:

PET/CT/Histopathology Reports
Blood Tests
Medicines
Bills/Invoices

Why? Doctors want quick access to specific information. If you mix blood tests with pharmacy bills, you create confusion. Classification gave structure.

📄 Step 3: Document Handling — Google Docs + Drive

For reports (PET/CT, Histopathology, Blood Tests), I set up a Google Docs creation and update flow:

Search Files & Folders Node → Look up if a document already exists for this category.

IF Node → If it exists → Update Document; else → Create New Document.

Why? Doctors often ask for longitudinal comparisons (e.g., “show me how her PET scans have evolved”). By updating existing documents instead of endlessly creating new ones, we built living documents with the full medical history.

💊 Step 4: Medicines Management — Structured Sheets

Medicine prices vary across pharmacies, and availability is tricky. Here I used:

Google Sheets → As a structured store for medicines.

Bright Data Scraper → To fetch real-time medicine prices and availability from GoodRx.

Custom Code Node → To dynamically generate drug URLs from sheet entries.

Why? When managing long-term prescriptions, cost matters. Automating price comparisons across pharmacies helped us save money and plan better.

📊 Step 5: Bills & Invoices — Organized Proofs

Bills and invoices were funneled into a Google Doc draft. Why a draft? Because sometimes these need manual edits before submission for insurance claims. By automating drafts, I reduced the grunt work while keeping flexibility.

🌐 Step 6: Bright Data + HTTP Requests — Real-Time Intelligence

Some data needed scraping beyond APIs. Initially, I struggled with datasets — I couldn’t find one directly in Bright Data’s dataset hub that fit my use case. After experimenting (and failing) to get prebuilt datasets working, I built a custom Bright Data collector with parser code. Then I connected it to n8n via the HTTP Request node to trigger scrapes dynamically and fetch results back into my workflow.

This was challenging but became the most powerful part of the system: real-time medicine price intelligence from the web. Also, the generate code feature was awesome and helped alot

📲 Step 7: Twilio Notifications — Closing the Loop

What good is all this if the caregiver (me) doesn’t know when new data arrives? I set up Twilio SMS/WhatsApp alerts at the end of each flow:

“New PET scan uploaded to Docs”

“Medicine price sheet updated”

“Invoice draft ready for insurance”

Why? Because healthcare is stressful enough. Getting real-time alerts meant I could focus on care, not file management.

Bright Data Verified Node

I used the Bright Data Verified Node to scrape structured drug pricing and pharmacy details from GoodRx. It dynamically triggered collectors, fetched clean JSON, and integrated the results into Google Sheets and Docs. This allowed real-time tracking of medicine prices, availability, and related insights within my caregiver workflow.

Journey

This project wasn’t about shiny tech — it was about reducing cognitive overload for caregivers. By chaining together Gmail, AI classifiers, Google Docs, Sheets, Bright Data scrapers, and Twilio alerts in n8n, I built my first AI agent and a personal healthcare automation system.

At first, it was frustrating to get Bright Data datasets connected — but that challenge pushed me to learn how to build my own collector and integrate it with HTTP nodes. That struggle turned into one of the most rewarding parts of this journey.

For my mother, it meant we always had answers when doctors asked for history. For me, it was a reminder that the best tech emerges when we solve problems close to home.