Simetrik

Hello, I’m Alejandro, I have 6 years of work experience (5 at Simetrik), and I would like to share a bit about my experience working at a startup.

It was 2018, and I was finishing my internship at Avianca. The first thing we were told when we arrived at El Dorado Airport, the operations team, was: “I don’t understand why they sent a group of 4 mechatronics engineers, let’s see what they start doing.” That’s how, with my fellow interns, we began discussing the best way to invest our next 6 months (beyond just sitting around warming up our positions).

Initial Context

I am a person who has lived with duality. My parents separated when I was two years old. This allowed me to always be immersed in different situations that required me to step out of my comfort zone. My father is from Sogamoso, a relatively small city compared to Bogotá, where he spent most of his time with my uncle and cousins playing. My mother is from Bogotá, where everything was chaotic, and I accompanied her to her job during her free time (a clothing store in Chapinero, where I later attended university).

My mother was always a very demanding but loving person. She dedicated herself to explaining why things were the way they were and always motivated me to try new things. “Try it, try it; if you don’t like it, that’s fine, but don’t tell me you don’t want to without trying.” My father, on the other hand, always instilled good manners, the importance of keeping one’s word, and the commitments made.

This, in one way or another, helped me develop certain skills, especially the desire to be in constant change and to try new, challenging, and interesting things (although it was difficult for my mother, she managed it).

After studying at an all-boys school in Bogotá, a co-ed school in Sogamoso, and then returning to Bogotá for university, the idea of interning at Avianca was almost like a dream. A renowned company with over 17,000 employees and countless destinations and daily flights represented an unimaginable development opportunity for me. That’s why I accepted, and why the blow was so hard when I received the phrase: “I don’t understand why they sent a group of 4 mechatronics engineers, let’s see what they start doing.”

Avianca

We reached an agreement where my three colleagues and I would propose solutions to the executives based on our knowledge, and if they liked them, we would start developing them, presenting progress and so on. The first thing we did was understand the business: we had to identify the main problems to present solutions to real issues. We spent a month attending the airport at different times and came to the following conclusions:

• Lack of real-time information: At any moment, the itinerary changed due to operational difficulties, forcing those in charge to constantly move to Avianca’s assigned gate (33) to obtain the new itinerary. This created, especially, dead times in the operation.
• Lost luggage: There were a lot of lost suitcases that required having a storage area for them so that, at some point, the user could claim them.

We proposed a chatbot to consume the itinerary in real-time as it arrived by email, allowing operators to check their corresponding flight through Telegram. We also proposed a mobile application where users, before traveling, could take photos of their luggage, including the flight number and name. Thus, if for any reason a suitcase ended up in the lost baggage area, the operator could take a photo of the suitcase and, using an AI and image detection algorithm, identify the possible owner.

We developed the products so that, in the end, the response was: “Very good, but it won’t be possible to implement because what already works is better left as it is.” This discouraged me a bit at the beginning; it went against everything I had imagined about the work world. Innovation was not encouraged, and there were obstacles for everything. However, I realized that the process of understanding a business, identifying its problems, and presenting solutions was an incredible challenge. The mere fact of going from an idea, researching a possible solution, and developing a product was something that excited me and continues to excite me to this day.

Interview at Simetrik

I finished my internship and started looking for a job. I didn’t know exactly where I wanted to work, but one thing was very clear: I didn’t want to work at a conventional company. A company that forced me to adhere to a schedule, imposed a rigid way of thinking, or limited me to explore and “break” things. For that, I understood that I couldn’t look for a job in a “normal” way. At that time, I was very into programming and machine learning, so in a Facebook group called “Machine Learning Colombia” I wrote a post:

“Hello, my name is Alejandro, I’m 23 years old, about to graduate in mechatronic engineering, and I have programming knowledge. I’m looking for my first job opportunity in the data field.” I received all kinds of comments, but the most important was when Heinz Sohm wrote to me directly: “Hello Alejo, we are building a product with some partners and we are looking for a profile like yours. If you want, I’ll tell my partner to write to you.” Of course, I agreed. Around 6 PM, Alejandro Casas called me and explained the problem: “We are building a reconciliation software for Rappi. They have a problem with the money coming in and going out. It’s a lot of information, and we’re figuring out how to solve it.” We talked for almost 3 hours, where Alejo asked for my opinion on what he was explaining, how I would solve it, or what ideas I had. Finally, he said: “Why don’t you come to Rappi, we’ll talk in person, meet the team, and decide what to do?”. I immediately said yes (I clarify that we didn’t set a time or anything, we just agreed that I would go talk to him).

After searching Google for Heinz and Alejandro, I came across this article: Ropeo seeks market in Ibagué and Armenia. It turned out that the people I had just spoken with already had a company and had received investment. This excited me; they were people involved in the entrepreneurial scene and were betting on creating innovative solutions to problems they believed had growth potential. At that time, although the startup boom wasn’t as prominent, there were already success stories like Rappi, Uber, and Platzi. The latter was where I always read about entrepreneurship, accelerators, and technological topics.

I woke up around 8, took the bus at 9, and by 10, upon arriving, I received a call from Alejandro: “Hey, I’ve been waiting for you for an hour and you haven’t shown up. If this is your attitude with your new job, we’re off to a bad start.” I was surprised because that wasn’t how things had been; however, I apologized and told him I was just a few blocks away. I arrived at Rappi’s offices on Calle 100 with 19. A building with a mustache logo, almost 8 floors high. I was surprised when Alejandro came down looking as relaxed as possible, a guy who didn’t look older than 28 (in my opinion). He introduced himself, took me to the third floor at Rappi where most of the rest of the team was: Heinz, Jesus, María Paula, and Duván. We went to the 7th floor to an independent room, and he said: “Tell me about yourself, what you like to do, what sets you apart from others. But don’t tell me where you studied, I’m not interested. I want you to talk from your experience and why you’d like to work with us.” This surprised me a bit, honestly. It was the first interview where things were so informal. Additionally, a lot of importance was placed (from my perspective) on who I was from my essence (experience), not from the material (place of study).

“I am Alejandro, I’m 23 years old, about to graduate in mechatronic engineering, I did a project focused on automated organic crops, and I’m looking for a company that helps me grow. I want to learn and I’m willing to do whatever is necessary. I like programming because I feel I can solve any problem with programming.” He said to me: “Alright, let’s do something: do you know how to use Angular?” (a frontend development framework). “No idea,” I said. “Perfect, then you’re going to make an app that allows a form to receive product recommendations, store them in a cloud database, and display them. When would you have it done?”, Alejo asked. “In a week,” I said without knowing. “Perfect, we’ll see you in a week,” Alejo said. “If you have nothing or can’t finish, don’t even think about coming,” he added at the end 😂. I said: “Just one thing, Alejo. I’m currently preparing to go study for a master’s in Germany. I’m working on my English and other things, the idea would be to travel in about 9 months.” Alejo said: “So don’t worry, you follow your plans, work with us, and if you decide to leave afterward, then you leave, and if you like it, then you stay. So no problem.”

I left that meeting very excited for several reasons:

1. I was facing a group of people who wanted to do interesting things. Young, with different knowledge, but very committed to what they were doing. With a passion for solving problems that, until then, was unknown to me.
2. They weren’t interested in where I studied, but in what I knew and my attitude. It’s something I hadn’t experienced elsewhere.
3. I greatly appreciated the fact of receiving reassurance regarding my plan to travel to Germany. I felt Alejandro so convinced of what he was doing that he was sure I would stay at Simetrik. And he might not have, but the fact that he still considered me for the opportunity encouraged me to do better.

Trial and First Days

I spent a week watching many tutorials on the internet about Angular and Firebase. I finished the application a day early and went to talk to Alejo. I showed him the application, he liked it, and he said: “Alright, can you start now?” I said yes, immediately. The initial arrangement was for two weeks to see how I fared with the team and then, if all went well, they would hire me directly. The plan was to support Duván as a Frontend developer. That’s when they introduced me to the team:

• Heinz: A kind, helpful, and very analytical person. He worked alongside Jesús on developing the reconciliation algorithm. They talked a lot about ghosts, pandas, and Python.
• Jesús: Heinz’s battle-hardened colleague. They had discussions that at that moment to me were “heated” to reach problem solutions. Generally, with Alejo. We got along well from the start thanks to video games.
• Alejandro Casas (Alejo): At that time, I felt he was the gear of the entire team. He did personnel selection, talked about the technological layer of the product, user experience, understood the user’s problem, and sold it. The visionary who, along with Santiago (Santi), would change the focus overnight in favor of the product’s timelines and deliverables.
• Duván: Frontend developer with whom I worked closely for a month (while we had a frontend that changed thanks to conversations with Santi and Alejo). His dedication and commitment were enviable. He was constantly growing, as if his hunger for knowledge could never be satisfied.
• Santiago (Santi): I met him almost a week after joining the team. At that time, I thought Alejo had a strong character until I met him. He is a person who does whatever is necessary to get things done. He can take any type of problem and focus efforts on the necessary time to deliver in the best way.
• María Paula: The team’s designer, who helped untangle concepts and clarify the focus that Alejo wanted for the product. Her design skills ultimately resulted in products aligned with user expectations.

I initially worked with Duván, with whom we were able to create the first version of Simetrik (actually, I only contributed a load bar, at most). A SaaS that allowed creating sources from different origins and viewing each source on a screen in a box, connecting them with arrows, and showing the reconciled percentage.

This was one of the most impactful moments for me personally and professionally. I was poorly seated, looking at code I didn’t understand, when Alejo approached and said: “Do you think that’s a position to work in?” and started asking me questions related to the code and data flow. I was able to respond as best as I could, but in the end, Alejo gave me feedback on commitment and how I should approach challenges to meet the level of commitment that Simetrik required. Every day I thought they were going to fire me, felt like I wasn’t performing, and that I couldn’t meet the level of passion and commitment. However, I wasn’t going to give up. I knew that to grow, I had to adjust many things and adapt to others.

One of the things I have tried to focus on lately and find important to continue growing as a person and as a collaborator in a company is not taking anything personally and understanding this flow of feedback as a transparent way to improve.

Everything was going well in the product development until we tried to connect the frontend with the backend. The reconciliation algorithm wasn’t working as expected, and the deadline was approaching.

One day, Alejo arrived, said we had to stop everything, and started a process that, in retrospect, is what I have learned the most from. Basically, everything that had been worked on for 2 months was halted, and it was decided to start from scratch. The priority was to upload Rappi’s sources to a database and generate valuable deliverables. That’s when a process of boards filled with ideas and solutions proposed by the team began to shape the deliverables.

Sebastián’s Arrival

I remember one day we were on TransMilenio with Duván, and he said: “I’m working at night with a machine, a top-notch expert. It seems Alejo wants him on the team, and the guy is flying.” I asked him what they were doing and so on. The case is that they were setting up the first reconciliation version in AWS’s EMR service. That’s when Alejo brought him in from Montería, and I met someone I consider one of my best mentors and great friend. That day, Santi approached to explain the project, what we were doing, and so on. He was surprised when Sebastián showed him that he was making improvements to the code. Clearly, Sebastián already knew what Simetrik was doing, how it was doing it, and was also improving it. From that day until today, there hasn’t been a day Sebastián isn’t involved in Simetrik’s technological decisions.

Although Sebastián is the engineering director now, I always see him working like the first day: with enthusiasm, dedication, effort, and a strong desire to conquer the world through his code.

Pivot to Data Engineer

Throughout the pivot process to this new version of Simetrik, Alejo called me one day and said: “Hey, I know we hired you to be a frontend developer, but now we need you to start downloading information, can you handle it? The same commitment is that at some point you automate the flow.” I said: “I will do whatever is necessary for Simetrik to grow, and if it means downloading data right now, I will do it.”

So, from that day, we started implementing the integration layer, the storage layer, and the reconciliation layer. My goal was to download information from different sources for Rappi: SFTP, Gmail, platforms, databases. We defined a data schema by country and by processor to store the data. I had to leave them in CSV format without headers and add the columns report_date and filename at the end. We had to download a lot of information, so Heinz helped me with information from Colombia’s database for several days of overnight work that I accidentally deleted (“my first job”). I thought everything would end at Simetrik for me, but fortunately, it didn’t.

I was downloading data manually, understanding the problem from all its angles. This allowed me, gradually and with a lot of research and team help, to start automating the flow.

1. I focused on emails. With the help of Zapier (a no-code integration tool), I sent the files to an S3 bucket based on how they arrived in the email subject.
2. I began transforming the data with Pandas and Python.
3. For SFTP, I came across WinSCP. A tool that allowed me to mirror data repositories locally (so I didn’t have to enter credentials one by one) with a .bat script.

I reached a point where I already had a script on my computer that, when executed, performed the entire flow that I initially had to do manually. The next step was to use Lambda functions. Here, and through events, I was able to make my first automation deployments to not depend on my computer. Everything was happening while Rappi was increasingly using the visuals being built in QuickSight with the reconciliation report tables.

As I progressed, I realized I wouldn’t always have the script on my computer. That’s when Jesús told me: “Upload it to Lambda.” At that moment, I didn’t even know what AWS was.

While this was happening, the product started generating value for Rappi’s financial team, and Simetrik began closing contracts with other clients like Grin and Coltanques. Additionally, we moved to a larger office and hired more people. One day I vividly remember in the Sancho office, Alejo told me: “Hey, this keeps growing, and you need support. We need someone who knows what you do and bring in more like you.” That day, I researched and found out he worked as a Data Engineer, and that’s when we started building the integrations team.

Simetrik continued growing and targeting interesting businesses, and mid-year, MercadoLibre bet on Simetrik. For the proof of concept, we had to fix some Prisma files. That’s when I started leading a team of 3 people with whom we developed ETL scripts. In November, I had the opportunity to accompany MercadoLibre’s implementation team and had to learn Django, plus a myriad of AWS services, to respond to their flow and processes. There, Jhonatan told me something that stayed with me forever: “You’re like a bloodhound waiting for its bone and won’t let go until you finish it.” We worked hard and implemented a flow that laid the groundwork for what would be the largest implementation to date at Simetrik.

During that process, over 70 people joined, and clearly, the Sancho office became too small. I lived incredible moments both personally and professionally there. I heard Alejo’s iconic word: “Fuego,” referring to an upcoming growth process, and we had to adapt to whatever came. “It’s a marathon, not a sprint,” he always said. People I remember a lot: Albeiro, Mario, Santi, Sebas, Alejo, Stephy, Mateo, Mauro, Zorro, Juanjo, Majo, Negra, Thali, Andrés, Dani, Luisca, Juancho, David, Jhonathan, Andrés, Sebas Ruiz, Diana, Jesu, Iván, Neifer.

That period also led us to today’s office, but we only enjoyed it for a few months before the pandemic (3 months).

One point I want to highlight here was when I deleted the Meli project on AWS in ELB because, to deploy, we had to upload a file without pressing the redeploy button since we didn’t have the permissions. It happened, and I spent half an hour telling Andrés that this was the end, and Jesús wasn’t going to forgive me. I had to write to him one day (his son’s birthday). The next day, we managed to fix it, and it became an anecdote. But I was really scared.

The good thing about this process is that it allowed me to understand the problem, tackle it in a “natural” way, and iterate little by little until reaching something robust that adapted to the business’s needs. It also made me realize that failing isn’t bad, but not learning from mistakes is. But above all, it taught me to enjoy the process and live one day at a time.

The First Integration Product

The process of running ETL in Lambdas worked at the beginning; but as Simetrik grew (in clients and in the number of people, reaching 90 by December 2019), the approach had to change to:

1. Control the processes (metadata).
2. Have a clear configuration layer (database).
3. Have a clear and independent processing layer (AWS Batch).
4. Scalability.
5. Observability.

That’s why we started with the first extraction product, which, to this day, processes over 7,000 data pipelines from the 90 clients Simetrik has. It was an interesting, challenging, and highly iterative process. I explained the idea to Jesús and Santi, who liked it a lot and supported me all the way. Additionally, Alejo guided the first project strategy. In parallel, we each went home due to the pandemic, and Simetrik gave me a loan (interest-free) to pay off my debt with Icetex (for my studies), which I am grateful for from the bottom of my heart (May 2020).

This period was one of significant growth. On one hand, we had to respond to client implementations (about 15 at the time), while thinking about the product, its infrastructure, development, and so on. It was a process of almost six months where we migrated the jobs to a process that met the criteria and more. I felt it allowed (and still does) fulfilling Simetrik’s and its clients’ needs (from the extraction point of view).

My Departure from Simetrik

By the end of 2020, I often felt overwhelmed with work and load. That’s when I had several conversations with Alejo, Santi, and Andrés about the future and how I thought we needed more support in integrations and allow ourselves to focus on operating, productizing, or developing, but not all three at the same time (we did it for six months, and with a lot of effort, everything turned out the best way, but the exhaustion was felt). The conclusion was that Simetrik still didn’t have a scalable process (from the reconciler’s perspective) and, therefore, the investment wouldn’t happen (at least for a while). Plus, I didn’t know or hadn’t had another work experience.

The reality is that this process was very tough for me. I left, and it wasn’t easy. I realized that being surrounded by intelligent people with the attitude to do whatever was necessary to fulfill and with values wasn’t common. Freddy Vega, at Simetrik Fest 2024, said: “One doesn’t realize they’re living a dream until they stop living it.” This phrase perfectly fits this process of my life.

I remember even talking with Andrés about the possibility of returning (a week after leaving). “Don’t rush,” Andrés said. “Go, get to know and see how life is elsewhere. We’ll talk in a year.” Indeed, that’s how it was. After a year away, Andrés called me: “Hey, what’s up?” We talked, and the conclusion was a proposal to return to Simetrik since they had achieved Series A, had the budget for technology, and wanted me to lead the integration efforts. I accepted with happiness and expectation; the idea was to start as a software architect.

Clearly, we first talked with Jesús and Sebastián about the challenge. The idea was to start generating efforts in the integration process for Simetrik’s NoCode. Honestly, working with them was already enough reason for me to return. Meanwhile, I had a failed business attempt: an application to manage financial control, and I realized I lacked many skills in team management and software development. That’s when I decided to pursue a master’s in software engineering at the University of the Andes, where Simetrik supported me throughout the entire degree with an amount they give to everyone who works there (500 USD annually).

My Experience as a Software Architect

When I returned to Simetrik, I found a team of over 150 people, mainly in technology. Additionally, I realized they had a complete team for product and another for operations. Having the time to focus on development was clearly the best surprise.

I worked closely with Sebastián and Jesús with a focus more on meeting quality attributes for infrastructure, software, and software architecture. A challenging and interesting process that even led me to lead the development process of the accounting translator and integrations. During this period, we released an MVP to convert reconciliation results into accounting entries and improved the extraction service to continue maintaining the operation.

It was interesting because, in parallel to improving my software development process skills thanks to the master’s, I was applying everything at Simetrik. Additionally, I had the opportunity to develop a solution that processes 250 million monthly JSON-type transactions (Meli site) in a fully managed AWS infrastructure with Kinesis, Firehose, S3, and Lambda.

At this point, Simetrik started seeing the possibility of using AI as a driver for different processes. With Daniel and Felipe, we won the first hackathon where we proposed a chatbot that used OpenAI’s API to address user needs through a prompt.

Here, I had the opportunity to start actively participating in the AWS community and later become a community builder through various blogs written (with Simetrik’s permission) and share my discoveries/developments with anyone who needed them.

One day I remember a lot was during a production release of the extraction project where we changed the duplicate detection algorithm and generated duplication across all Simetrik sources. We had to work overnight for two consecutive days. It was part of the growth but had a huge impact on clients.

Engineering Manager

Early 2024. Simetrik receives Series B from Goldman Sachs. The objective for integrations was to start opening its doors so that different processes could send information to the ecosystem. Sebastián and Jesús began building what are now known as the verticals. Teams focused on their domain but with a common goal: to take Simetrik global. Currently, I am leading a team of 20 people, and by the end of the year, we hope to continue building features to meet Simetrik’s objectives (from integrations).

Working with Valen, Marce, Dani, Tania, Aura, JuanDa, Lucho, Jorge, Diego, Paul, Fabián, Kevin, Paladines, Jose, Andrew, Juli, Aleja, Santi, Gaby, Mica, Jesús, Grace, Sebastián, Andy H, David, Néstor in integrations is one of the nicest experiences I’ve had, and I want to continue maintaining it.

To date, we have four extraction versions, and the first no-code integrations version will be released with the ability to manage data exports and with new services on the horizon (even with AI).

The process of being a manager is not easy, especially because it involves being transparent, knowing how to communicate, and having people skills. In the end, one succeeds through the team’s success. Facing difficult conversations (which clearly you don’t want to have) and, above all, having clear expectations from both sides, I feel, are key to properly focusing each person’s efforts.

This challenge has also been interesting; it allows me to see the project with a more general focus without losing sight of the particular. Additionally, it pushes me to step out of my comfort zone not only in soft skills but in strategies from the perspective of Infrastructure, architecture, frontend, QA, and backend.

I’m very excited to see not only the company’s growth (360 people from 17 countries to date, with 90 clients around the world) but also the growth of those who have been part of it. Being surrounded by people with the same values, focus, and desire to conquer the world is, in the end, what makes everything up to this point worthwhile. We’ll see how far this rocket takes us.

I write this first to thank Alejo and Santi for the opportunity. Their focus and commitment to Simetrik and those of us who work there have a multiplicative effect in many areas. Also, in a personal way, to express that, although the process of working at a startup has many challenges, it can be very enriching and provide, as in my case, a desire to continue growing to meet what my colleagues and Simetrik need. Thanks to those who have been part of this journey and to those who help me keep growing.

Alejofig.