This first-person essay, narrated through a conversation with Manoj Tumu, presents the journey of a 23-year-old machine learning engineer currently working at Meta’s Menlo Park campus. In order to ensure accessibility and precision, the narrative has been edited for both conciseness and clarity, and Business Insider has independently verified both Tumu’s employment status and his compensation package.
In recent years, the field of machine learning has transitioned from being a specialized discipline reserved primarily for academic researchers and niche industry experts into a widely recognized cornerstone of modern technology. The release of models such as ChatGPT in 2022 accelerated this transformation dramatically, bringing artificial intelligence into the public domain at an unprecedented scale. As innovative AI-powered tools began to capture the world’s attention, countless aspiring professionals sought to enter the industry, making it one of the most intensely competitive career paths.
Tumu’s educational trajectory was unconventional yet highly strategic. Because he had already accumulated substantial college credits through high school coursework, he was able to complete his undergraduate studies in only a single year. Rather than pausing after graduation, he seamlessly transitioned to working full time as an engineer while simultaneously pursuing a master’s degree in artificial intelligence, an experience that allowed him to gain both theoretical knowledge and applied expertise in parallel. Upon completion of his graduate studies, he secured a highly coveted role at Amazon, contributing as a machine learning software engineer for nine months. While he valued the depth of experience he gained there, he increasingly found himself drawn to the more ambitious and exploratory projects being developed at Meta. Motivated by curiosity and a desire for intellectual challenge, he submitted applications both through LinkedIn and directly via the company’s careers website. The opportunity he sought materialized in mid-2023, when he left Amazon to accept an offer at Meta with total annual compensation exceeding $400,000. For Tumu, the decision to accept was immediate and instinctive, as the scope of the work aligned closely with both his professional goals and personal enthusiasm for cutting-edge machine learning.
Reflecting on the nature of his current role, he describes how titles in the machine learning ecosystem can vary substantially between companies. The same kind of position might be described as an applied scientist, research scientist, software engineer, or machine learning engineer, depending on organizational culture and departmental objectives. At Meta, his official designation is “machine learning software engineer” within the advertising research division. This function combines elements of hands-on software implementation with a strong emphasis on research-driven innovation. A central responsibility of his job involves carefully analyzing newly published research papers and assessing whether emerging models can be integrated into Meta’s systems to ensure that the company remains at the forefront of technological advancement.
When commenting on the trajectory of the field itself, Tumu observes that machine learning has undergone a profound paradigm shift. In earlier years, many professionals relied heavily on classical techniques, which required human engineers to design and select the features they believed would best represent the raw data. Such strategies placed a premium on human intuition and statistical craftsmanship. In contrast, today’s research landscape is dominated by deep learning, a methodology that employs artificial neural networks capable of automatically deriving highly complex feature representations directly from raw information. Deep learning has proven remarkably powerful, and because of its successes, interest has grown not only among traditional technology companies but also among organizations in healthcare, finance, retail, and virtually every sector of the economy. These new applications have fueled a surge in both investment and hiring within machine learning.
For individuals aiming to launch a career in AI, Tumu stresses the importance of obtaining professional experience as early as possible. In his view, one of the strongest differentiators when reviewing résumés is whether candidates have completed internships. Even short-term or modestly paid experiences can play a decisive role when applying to competitive positions, as they demonstrate real-world competence that academic coursework or side projects do not always fully convey. While many applicants tend to emphasize programming languages or student projects, Tumu recommends prioritizing internships or employment opportunities, since professional responsibilities often carry more weight with recruiters at major technology companies.
He explains that his own résumé evolved over time: while early versions included detailed descriptions of personal projects, he eventually removed these once he had accrued several years of industry experience. At that stage, concrete professional achievements—such as his engineering contributions at Amazon—were far more compelling than student exercises. His advice is straightforward: projects remain useful as a supplement when one is just starting out, but once meaningful experience is on the résumé, it is appropriate to shift the focus entirely toward work history and impact.
Recruitment processes at major technology companies tend to follow a standardized yet rigorous format. Tumu recounts that Meta’s process resembled Amazon’s in many respects, involving an initial screening stage followed by four to six additional rounds of interviews. The assessments measured proficiency in programming, depth of knowledge in machine learning, and situational judgment via behavioral questions. From start to finish, the process took approximately six weeks. Despite its intensity, Tumu recalls it as one of the smoothest interview experiences of his career.
He cautions against one common mistake: underestimating the importance of behavioral interviews. Many candidates arrive confident in their technical ability but “wing” the behavioral component, assuming it requires little preparation. Tumu, however, took a different approach. He constructed a highly detailed document filled with carefully drafted stories, each tailored to potential behavioral questions. At Amazon, he studied the company’s leadership principles and aligned his responses with them; similarly, at Meta, he analyzed the company’s published core values and prepared narratives that explicitly demonstrated how his experiences embodied those ideals. This meticulous preparation allowed him to convey not only his technical competence but also his cultural fit.
Another lesson he shares involves compensation and career prioritization. At the beginning of one’s career, he argues, it is a mistake to focus excessively on salary. He personally chose a lower-paying machine learning contract role immediately after finishing his undergraduate studies, rather than pursuing potentially better-paid general software engineering opportunities. Although the compensation seemed modest at the time, that decision ultimately provided a crucial foundation in machine learning that made him more competitive for later opportunities at Amazon and, eventually, Meta. His advice for students is therefore to apply broadly for internships—even those with lower pay—because the long-term career growth such experiences enable far exceeds the short-term financial tradeoffs.
Ultimately, Manoj Tumu’s story demonstrates how an early focus on real-world experience, combined with resilience, curiosity, and thorough preparation, can accelerate one’s journey into the highest echelons of the technology industry. For anyone aspiring to break into AI and machine learning, his path serves as proof that with deliberate choices and disciplined effort, extraordinary opportunities can be within reach.
Sourse: https://www.businessinsider.com/quit-amazon-took-meta-ai-offer-how-to-land-job-2025-8
