Can you brief us about your scope of work during this internship?
I worked on two major projects during my time at ANB. One of them was improving HVAC nameplate extraction accuracy through both heuristics and machine learning models. This also involved working on the infrastructure to support these ML models. Another project was related to intelligent data extraction from various forms used by energy efficiency companies.
What were your key roles & responsibilities and how did you work towards achieving them?
While my responsibilities differed from project to project, I applied the same fundamental principles toward any task I was faced with. The basic workflow went from planning or outlining high-level objectives to executing those plans and applying programming best practices to the job at hand. Regarding my roles, I filled the role of a software engineer and was often given complete autonomy over my work. For example, I was told to improve HVAC extractions but not given a specific way or method I was supposed to use to do so. Coming up with techniques and solutions on my own was challenging but very rewarding.
Among your many tasks, what is the one that you liked working on the most?
I enjoyed working on improving the HVAC nameplate extractions because even though my work on this project spanned many weeks and involved multiple approaches, there was always a clear goal of increasing our reliability and accuracy. I was also involved in not only improving the extraction algorithms and results but also in the infrastructure related to them. Getting to develop all facets of the project was an extremely valuable learning experience.
What was the biggest challenge you came across and how did you overcome it?
Prior to my time at ANB, I did not have any workplace experience with machine learning technology. So, when I was tasked with developing an ML solution for HVAC nameplates, I was intimidated and sceptical about whether I could deliver. Fortunately, I had the help of more senior developers at ANB, who pointed me in the right direction. Using their help to drive my research, experimentation, and development, I successfully created a pipeline to train a model that would predict an HVAC unit’s manufacturer given a picture of the nameplate. Integrating this ML solution into ANB’s ecosystem also proved to be a challenge. However, with the help of my mentors, I got my model hooked up to AWS and into production without any major headaches.
What new skillsets and/or knowledge did you gain during your time with ANB?
Adding features and debugging code in ANB’s large codebase strengthened my skills and was a nice contrast to the more intimate personal or class projects that I was most accustomed to. Additionally, I heavily utilized Python for many purposes, including implementing algorithms, scripting, and building a REST API. These experiences have greatly bolstered my familiarity and competency with Python and programming in general.
What do you feel are the highlights of working at a ‘smaller’ tech company?
It was nice to have a greater individual impact and to practice rapid development. For example, I was given more autonomy during my work with HVAC nameplates, which allowed me to work quickly and efficiently. In a larger company, implementing my ideas may have been more difficult since they typically have more red tape and protocol that everything needs to go through before work can begin. This limits flexibility and rapid iteration. I also enjoyed being in frequent communication and working with leaders at ANB, which would be impractical at giant corporations.
What are the key takeaways from your exposure to the energy efficiency industry?Since I have never been exposed to the energy efficiency industry significantly before this internship, it was interesting to get some insights from this sector that I was unfamiliar with. Getting to tackle different challenges such as extracting form data and predicting manufacturers from nameplates forced me to shift my thinking and put myself in the shoes of someone in the industry. With this experience, I now feel like understanding the client’s perspective in any future endeavours will be easier, no matter the industry.
How did you find our company culture and the support you received from your colleagues?
Reaching out for support from various team members was greatly encouraged and benefitted me significantly. Everyone I talked to was happy to help me get set up with new technology, walk me through existing code, or get me unstuck while working on a project. I always felt welcome.
How easy/challenging was it to collaborate remotely with COVID-19 restrictions in place?
Remote collaboration was surprisingly smooth with COVID-19 restrictions. Video calls and sharing screens were good alternatives for in-person work. I worked with the ANB Houston office, which works with the Chennai office in India. Overall, the communication process was completely seamless.
How much did the internship match your expectations?
My time at ANB exceeded my expectations, and I would highly recommend the experience to anyone interested in a fulfilling and educational internship.
How much do you feel your internship will help you in your career going forward?
I feel like the internship taught me a lot about working on a software team, and the lessons I’ve learned will aid me greatly.
