Question 1: Describe a time when you improved a testing process.

Answer: Situation: At Spark Cloud Studio, I joined as a Software Test Engineer Intern and was
quickly immersed in a project aimed at enhancing the testing workflows for both web and native
applications. Task: My primary goal was to identify and rectify bugs, while also boosting the
overall efficiency and quality of our software releases. Action: As I am quite Inventive and
Simplify many details, I implemented automated testing workflows using Playwright and
meticulously developed comprehensive test cases with Selenium and JIRA. I also integrated
continuous integration/continuous deployment (CI/CD) pipelines using Docker and AWS to
streamline the entire process. Result: These enhancements led to a significant decrease in bugs
by up to 86% and increased the testing efficiency by 40%, ensuring that I consistently delivered
high-quality software. This not only improved my delivery timelines but also enhanced customer
satisfaction with the reliability of company’s products.

Question 2: Tell me about a time when you used innovative technology to solve a
problem.

Answer: Situation: During my internship at Hexagon AB in India, I was tasked with enhancing
the data retrieval system used for spatial and string-based queries, critical for the defense team's
operations. Task: The specific challenge was to integrate a new feature to improve the system’s
efficiency and accuracy. Action: Demonstrating Learn and Be Curious, I developed an
interactive GUI using Python and LuciadLightSpeed software, which enabled real-time
simulation of missile trajectories. This project required me to acquire new skills rapidly and
apply them effectively to meet the project's demands. Result: This innovative solution improved
the search efficiency by 20% and was praised for its accuracy and speed, significantly reducing
the navigation time within our systems. My ability to leverage cutting-edge technology provided
substantial value to the client, enhancing our company's reputation for innovation.

Question 3: Can you discuss a time when you had to collaborate with a diverse
team to deliver a project?

Answer: Situation: At Spark Cloud Studio, I was part of a project that required close
collaboration with cross-functional teams to develop and test software efficiently. Task: Our
objective was to enhance project workflows and improve communication between the
development and testing teams. Action: Embracing Amazon's Dive Deep and Earn Trust
principles, I actively engaged in both front-end and back-end development, utilizing Svelte on
the front-end and Vercel along with Docker for back-end functionalities. I also focused on
optimizing the deployment processes and implementing best practices in collaboration with team
members. Result: This collaborative approach significantly improved our software testing and
development practices, streamlining workflows and enhancing the effectiveness of our
communication and progress tracking. The project was completed on schedule, under budget,
and with superior quality, leading to high client satisfaction.

Question 4: Describe a project where you dealt with ambiguity.

Answer: Situation: At Insigma Technologies, I was tasked with developing an AI system to
assess mask compliance during the uncertain early days of the COVID-19 pandemic. Task: The
project was fraught with ambiguity, dealing with various mask types and differing compliance
standards across multiple locations. Action: Applying Frugality and Bias for Action, I designed
and deployed a Convolutional Neural Network (CNN) that integrated with live security camera
feeds to analyze mask compliance in real-time. This involved managing extensive data
collection, which included over 2000 images of different mask types, performing data
augmentation, and tuning hyperparameters to optimize the model’s performance. Result: The
system effectively reduced manual verification time by 75% and provided a reliable, automated
compliance reporting tool across more than 50 locations, significantly enhancing the
organization's health safety measures during a critical period. This project not only supported
public health measures but also demonstrated the potential of AI in enhancing safety protocols
during critical times.

Question 1: Tell me about a time when you had to make a decision with
incomplete data.

Answer: Situation: At Insigma Technologies, I faced a significant challenge when tasked with
developing an AI system to monitor mask compliance during the early stages of the pandemic,
where data was notably sparse. Task: My task was to enhance the accuracy of our model to
ensure reliable compliance across multiple regions, despite the glaring data gaps. Action:
Embracing Amazon's Bias for Action, I took proactive steps to extrapolate from the limited data
available and consulted with health experts to estimate missing regional standards. I integrated
these insights into our model's algorithm to refine its predictive capabilities. Result: This
approach proved successful as the system was deployed across 10 locations, achieving a 15%
improvement in compliance accuracy over our initial models. This not only demonstrated my
ability to operate effectively with incomplete information but also highlighted my commitment
to delivering results in challenging conditions.

Question 4: Give an example of an initiative you undertook that benefited the

company but wasn’t within your direct responsibilities.

Answer: Situation: At Spark Cloud Studio, I noticed that our software deployment processes
were critically outdated, affecting the overall efficiency of operations. Task: Although updating
these processes was beyond my direct responsibilities, I recognized the broader impact a revamp
could have on the company. Action: Demonstrating Ownership, I took it upon myself to
research and implement modern CI/CD pipelines using Docker and AWS. This initiative
required me to not only acquire new technical skills but also to lead the integration effort across
our development teams. Result: The new deployment processes improved efficiency by 40%,
reducing deployment times by an average of 6 hours per cycle, which significantly enhanced our
project lifecycle and minimized downtime. This initiative not only boosted productivity but also
fostered a culture of continuous improvement within the company.

Question 5: Describe a time when you took on work outside your comfort area.
Answer: Situation: At Hexagon AB, an urgent requirement arose for an interactive GUI on a
project designed for missile trajectory simulations—a skill outside my backend development
expertise. Task: I was tasked with quickly mastering frontend development to meet the project's
critical deadline. Action: Leveraging Learn and Be Curious, I immersed myself in Python GUI
development using LuciadLightSpeed software. I dedicated extra hours to learning this new skill
and collaborated closely with the UI/UX team to ensure the functionality aligned with user
needs. Result: The GUI I developed was implemented successfully, enhancing operational
efficiency by 30% for the defense team's missile trajectory simulations. My willingness to step
out of my comfort zone not only contributed to the project's success but also expanded my skill
set, demonstrating my adaptability and eagerness to tackle new challenges.

6 & 7. Tell me about a time when you worked against tight deadlines / Delivered
an important project under a tight deadline.

Situation: During a high-intensity hackathon at Arizona State University, I led a project titled
"Event Reasoning with Explicit Time", which focused on enhancing temporal reasoning
capabilities in natural language processing.

Task: The challenge was to improve the performance of large language models (LLMs) on
nuanced temporal tasks using the innovative approach of instruction tuning. Utilizing benchmark
datasets like TRAM and TimeBench, my task was to reformat these into instruction-response
pairs, tune the Llama-3.2-1B-Instruct model from Hugging Face, and significantly boost the
model's temporal reasoning accuracy—all within the strict 48-hour deadline of the hackathon.

Action: As the team leader, I spearheaded the project by coordinating the efforts of my
teammates and distributing tasks according to each member's strengths. I personally took charge
of the most critical aspects:

 Data Preparation: I reformatted the datasets into instruction-response pairs, ensuring

they were primed for effective model training.
 Model Tuning: I oversaw the instruction tuning process, making sophisticated
adjustments to model parameters to better assimilate and process temporal information.
 Resource Management: I facilitated the use of cutting-edge resources like ASU-SOL
and Google Colab's NVIDIA A100 GPUs, ensuring efficient and effective training
sessions.

Result: Our efforts were highly successful. The post-tuning performance of our model showed a
significant improvement, with Exact Match Accuracy increasing from 1.95% to 4.22%. This was
complemented by substantial gains in other key metrics such as Average Levenshtein Similarity
and ROUGE scores. My leadership and strategic execution under the tight deadline not only led
to these remarkable improvements but also contributed to advancing our understanding of
temporal reasoning in NLP.

Impact: This project not only demonstrated our model's enhanced ability to comprehend and
reason about temporal data accurately but also set the stage for further research and application
in real-world scenarios. The project was a collaborative effort, and I am proud to have led a team
that included talented peers under the mentorship of Professor Chitta Baral. Our work has
opened new avenues for research and has potential applications in various technology sectors.

Question 9: Tell me about a time you dealt with a difficult customer or

stakeholders.

Answer: Situation: At Spark Cloud Studio, I was challenged by significant client reservations
about new features we were proposing, which threatened to derail the ongoing project. Task: My
responsibility was to address the client's concerns in a manner that not only alleviated their
doubts but also kept the project on track without compromising on quality or deadlines. Action:
Implementing Amazon's principle of Earn Trust, I initiated several rounds of detailed,
transparent discussions with the client, offering clear demonstrations of the new features and
providing data-driven insights into how these would benefit them. I supplemented these
discussions with additional support, adjusting project specifics to better align with the client’s
expectations and needs. Result: My focused efforts in building trust and actively engaging with
the client paid off. Not only were their concerns thoroughly addressed, but we also retained the
client, preserving an annual revenue of approximately $200,000. This success further solidified
our reputation as a customer-centric organization that values client input and adapts solutions to
meet their needs.

Question 10: Share an example of what you would do if faced with adversity or a
business challenge.

Answer: Situation: During my remote internship at Insigma Technologies, the onset of the
COVID-19 pandemic introduced significant operational disruptions as we were transitioned to
work remotely. Task: Faced with the adversity of maintaining productivity and project timelines,
I needed to quickly find solutions to adapt our workflow to a remote setting without sacrificing
project quality. Action: Embodying Amazon's Invent and Simplify principle, I led the initiative
to implement robust remote work tools and strategies. I evaluated several digital collaboration
platforms before selecting the most suitable ones, set up remote access to our development
environments, and trained the team on best practices for remote work. Result: My proactive
measures ensured that our team not only adapted quickly to the new remote work environment
but also met all project deadlines. We maintained a 100% project delivery rate, which preserved
client satisfaction and upheld our contractual commitments, demonstrating resilience and a
commitment to operational excellence even under challenging circumstances.

Question 11: Please provide an example of a time when you exceeded customer
expectations.

Answer: Situation: In a project at my university, we aimed to elevate the user experience with
an AI-driven music recommendation system integrated with Spotify. Task: The goal was to
provide users with a seamless and highly interactive music discovery experience that went
beyond traditional streaming functionalities. Action: Leveraging Amazon's Customer
Obsession, I took the initiative to integrate the Spotify Web Playback SDK into our application,
which allowed users to directly control playback within the app. This feature was complemented
with direct song previews and customized playlists based on user behavior and preferences.
Result: The integration significantly enhanced user interaction with the platform, leading to a
user satisfaction rate soaring over 95%. The system maintained a 99% uptime even under peak
load conditions, contributing to a 20% increase in user engagement. This innovation not only met
but exceeded customer expectations, showcasing our commitment to enhancing user satisfaction
through continuous innovation and customer-focused features.