Walking into my first day at Vatico, I expected to spend most of my time getting acquainted with tools and processes. What I did not expect was how quickly the day would challenge me to think not just about what I was doing, but why and how well I was doing it.
As the day draws to a close, I want to capture three things that stood out to me: what I learnt, what I appreciated about Vatico, and what I think can be done better.
The Art of Writing a Good Ticket
Before today, I had never given much thought to how a task is communicated. A ticket, to me, was simply a to-do item. That assumption was quickly dismantled.
For interns at Vatico, the expectation is clear: this template is non-negotiable and must be followed strictly — Background sets the context, Objective defines the why, Action describes the what, and Deliverable specifies the concrete output expected.
When starting out as an intern, it isn’t immediately obvious what a well-crafted ticket should look like. Following this template strictly is a mandatory discipline to practice effective ticket writing—it is to ensure that we explicitly articulate what we are working on, why we are doing it, and how we will accomplish it, while ensuring team members receive a solid high-level overview of the task.
However, the broader lesson is that beyond internship expectations, you don’t necessarily have to rely on these rigid sections forever. A clean structure alone does not make a ticket work — at its core, the primary goal of any ticket is to communicate your task and problem clearly to someone else. Without deep contextual clarity, even the best-formatted template falls short.
It struck me that providing clear context is not merely administrative tidiness — it is a form of respect for the reader’s time and clarity of thought. Without it, the recipient is forced to interpret and guess, introducing error and inefficiency before the work has even begun.
Dashboard Wireframing and Data Storytelling
I was also introduced to the principles of dashboard wireframing, which was entirely a new territory for me. I learnt to think in terms of a container layout rather than filling space reactively with visuals. The goal: minimise scrolling and present everything meaningfully within a single view — a discipline that forces you to prioritise what truly matters.
Equally illuminating was the framework for data storytelling, built around four key questions:
Total volume over time
Breakdown by channel
Rate or average
Incentive or driver
These four questions gave me a mental scaffolding for approaching any dataset with intention, rather than querying
aimlessly. On the technical side, I deepened my understanding of SQL concepts — CASE WHEN, CTEs with
WITH, and Window Functions — and gained clarity on the subtle but important difference between using
= and BETWEEN when filtering by dates.
Finally, I gained a deeper understanding of why ELT fits Vatico compared to traditional ETL (Extract, Transform, Load). Data architecture choices are always relative to an organization’s specific factors — such as data culture, database size, and pipeline complexity. At Vatico, the primary reason ELT fits our architecture is transparency. By extracting and loading raw data into the warehouse first before applying transformations downstream, we preserve the untouched data as a single source of truth. This makes inspecting raw data and deriving true original values simple and direct, avoiding the extra engineering overhead needed under ETL to reconstruct pre-transformation history.
| Dimension | ETL Architecture | ELT Fit for Vatico |
|---|---|---|
| Execution Sequence | Data transformed before loading into the data warehouse | Raw data loaded first; transformations run downstream inside the warehouse |
| Transparency & Source of Truth | Preserving or inspecting original raw history requires extra pipeline engineering | Native Transparency: Retains untouched raw data as a single source of truth for easy inspection and re-derivation |
What I Liked About My First Day at Vatico
What genuinely impressed me was the intentionality behind every practice. Nothing felt arbitrary. From the way tickets are written to the way dashboards are designed, there is a clear philosophy in how things are done – do things in a way that is scalable, maintainable, and built to last.
The DA team’s development pipeline is a perfect illustration of this thinking in systems:
I also appreciated that on my very first day, I was given homework that pushed me to reflect, not just absorb. Being asked to compare a good ticket to a bad one, and to articulate the difference between an Objective and a Deliverable, forced me to internalise the lesson rather than merely observe it.
Beyond that, a lot of effort was placed into ensuring interns are given thorough documentation about their job roles — uncommon, and incredibly helpful for understanding the company’s conventions and what to expect when working at Vatico. This alone saves countless hours, as standards, best coding practices, and workflows are already well defined, leaving little room for ambiguity.
Lastly, having a Telegram bot to field questions is extremely useful — and it speaks to how Vatico genuinely cares that interns are learning, not just left in the dark to figure things out on their own.
What Can Be Improved
- Broken resource links — Several links in the BVS data analytics materials return 404 errors, which interrupts the learning flow for a newcomer trying to explore independently.
- Outdated tool references — The slides still reference pgAdmin, even though it is no longer in use. A minor inconsistency, but one that causes confusion when reality does not match the documentation.
- Nice-to-have: a DocHub — A centralised documentation hub (similar to Python’s official docs) for important standards and conventions would be a great addition — though the current slides are already sufficient and helpful.
Proposed Solution: Confluence + AI — While adopting a Telegram bot for fielding questions is a great step forward, maintaining up-to-date documentation on Confluence remains a challenge as processes evolve. A key consideration moving forward is pairing Confluence with AI to automatically keep our documentation updated with proper, rich context — serving as the overarching solution for all the above improvements.
Beyond documentation, there should be a clear expectation set within Vatico: call out the problem early and work it out along the way, allowing the team to gradually address the problem rather than leaving it unnoticed.
Day one. Small steps — but already, a much bigger picture of what thoughtful, intentional work can look like.

