Career Development Path for Algorithm Engineers

This article is based on a talk by Tan Menglong, the former Director of Algorithms at ZhiZhuan, summarizing his career development experience and advice for algorithm engineers.

1. Career stages and experiences: Tan started as a senior R&D engineer in a search ranking team at a major internet company, describing this period as his "career university." He then moved to mobile search as a technical expert, calling it "first entry into the field." Later he joined a used‑car e‑commerce startup as a technical director, gaining complex business, team‑management, and commercial sense. Finally he became Algorithm Director at a second‑hand goods platform, focusing on large‑scale AI practice, team building, and business system construction.

2. From algorithm engineer to director: The required capabilities at each level are listed:

Algorithm Engineer: solid theory, implementation ability, communication.

Algorithm Expert: domain understanding, problem decomposition, technical leadership, foresight.

Algorithm Manager: line management, cross‑role collaboration, upward communication, mindset shift.

Algorithm Director: second‑line management, team formation, system building, business sense.

3. Choosing opportunities in the AI era: Tan explains the technology adoption curve (breakthrough → scalable application → commercialization) and advises self‑recognition, understanding industry positioning, setting primary and secondary goals, and balancing compromises.

4. Practical Q&A discussion:

• For small data scenarios, collect data aggressively and consider association‑rule algorithms, but remember recommendation is an end‑to‑end process linking resources and customers.

• When balancing data, backend, and algorithm work, first define the desired technical role (research‑oriented, implementation‑oriented, or leadership) and align tasks accordingly.

• Regarding graduate studies, clarify the purpose, identify skill gaps, and assess whether work experience can fill them.

• Effective learning channels include targeted books, paid courses, and papers, always tying learning to practical application.

• To persuade a boss about a new recommendation algorithm, understand external trends, propose strategic designs, and demonstrate value through real‑world results.

The article concludes with a reminder that continuous self‑assessment, leveraging diverse information sources, and applying cross‑domain thinking are essential for sustained career growth.