搜索结果：LinkedIn

Job understanding is critical to LinkedIn's mission of connecting talent with opportunity. This task involves transforming unstructured and noisy job postings into standardized or derived job attributes that power numerous LinkedIn products. However, building a scalable, cost-efficient, and high-performing job understanding system remains challenging. In this paper, we present a unified semantic modeling framework powered by a small language model (SLM) to address the challenges. We begin by fine-tuning an open-source SLM using a suite of carefully curated synthetic tasks augmented with reasoning traces. These tasks jointly target taxonomy-guided classification and taxonomy-agnostic entity extraction. This allows the resulting model to acquire robust zero-shot generalization for job understanding in structured and unstructured contexts. Building upon this foundation, we introduce a multi-adapter architecture with attribute grouping to facilitate efficient task-specific adaptation while streamlining model management across diverse downstream attributes. Offline evaluations and online A/B tests demonstrate significant performance improvement while reducing operational complexity. Our

LinkedIn is the largest professional network in the world. As such, it can serve to build bridges between practitioners, whose daily work is software engineering (SE), and researchers, who work to advance the field of software engineering. We know that such a metaphorical bridge exists: SE research findings are sometimes shared on LinkedIn and commented on by software practitioners. Yet, we do not know what state the bridge is in. Therefore, we quantitatively and qualitatively investigate how SE practitioners and researchers approach each other via public LinkedIn discussions and what both sides can contribute to effective science communication. We found that a considerable proportion of LinkedIn posts on SE research are written by people who are not the paper authors (39%). Further, 71% of all comments in our dataset are from people in the industry, but only every second post receives at least one comment at all. Based on our findings, we formulate concrete advice for researchers and practitioners to make sharing new research findings on LinkedIn more fruitful.

Production analytics products often depend on reusable methodologies: multi-step definitions such as headcount growth, top-skill growth, or differentially-private impression distributions. Although these methodologies define business-critical numbers, they are commonly implemented as imperative glue around OLAP queries, service calls, joins, transformations, and conditional logic. As a result, teams duplicate orchestration code, definitions drift across products, and methodologies are difficult to test or analyze. We present QDAG, a production system at LinkedIn that represents an analytics methodology as a declarative directed acyclic graph of typed steps. Nodes may execute Apache Pinot queries, downstream service calls, in-memory SQLite joins, jq transformations, conditionals, differentially-private aggregations, or calls to other QDAGs. The engine evaluates graphs demand-driven, memoized, pruned, and parallelized in the per-request analytics mid-tier. QDAG is deployed across more than 500 hosts and over 100 production use cases, adding roughly 10 ms median orchestration overhead and under 50 ms at the 99th percentile. Our experience shows that making methodologies declarative im

This study explores the relationship between LinkedIn profile characteristics and professional success, focusing on the indicators of promotions, follower count, and career progression rate. By leveraging a dataset of over 62,000 anonymized LinkedIn profiles, we developed predictive models using machine learning techniques to identify the most influential factors driving professional success. Results indicate that while promotions are highly predictable, follower growth exhibits greater complexity. This research provides actionable insights for professionals seeking to optimize their LinkedIn presence and career strategies.

Real-time OLAP datastores are critical infrastructure for modern enterprises, powering interactive analytics on petabyte-scale datasets with subsecond latency requirements. As these systems become integral to service architectures, maintaining strict SLAs under failures, load spikes, and cluster changes is as important as raw performance. We present a set of resiliency mechanisms developed for Apache Pinot at LinkedIn, applicable to modern OLAP systems broadly. We introduce Query Workload Isolation (QWI), which provides workload-level CPU and memory budgeting across Pinot's broker and server tiers via fine-grained resource accounting and sub-millisecond enforcement, delivering predictable tail latency and fairness with under 1% overhead. We present Impact-Free Rebalancing for SLA-safe data movement during routine operations (e.g., upgrades, scale-out, and recovery), and Maintenance Zone Awareness to place replicas across fault domains and mitigate correlated failures. We also describe Adaptive Server Selection, which routes queries using real-time load and performance signals to avoid slow or failing nodes while preserving balanced utilization. Together, these mechanisms form a hol

Notification recommendation systems are critical to driving user engagement on professional platforms like LinkedIn. Designing such systems involves integrating heterogeneous signals across domains, capturing temporal dynamics, and optimizing for multiple, often competing, objectives. Graph Neural Networks (GNNs) provide a powerful framework for modeling complex interactions in such environments. In this paper, we present a cross-domain GNN-based system deployed at LinkedIn that unifies user, content, and activity signals into a single, large-scale graph. By training on this cross-domain structure, our model significantly outperforms single-domain baselines on key tasks, including click-through rate (CTR) prediction and professional engagement. We introduce architectural innovations including temporal modeling and multi-task learning, which further enhance performance. Deployed in LinkedIn's notification system, our approach led to a 0.10% lift in weekly active users and a 0.62% improvement in CTR. We detail our graph construction process, model design, training pipeline, and both offline and online evaluations. Our work demonstrates the scalability and effectiveness of cross-domai

In large scale recommendation systems like the LinkedIn Feed, the retrieval stage is critical for narrowing hundreds of millions of potential candidates to a manageable subset for ranking. LinkedIn's Feed serves suggested content from outside of the member's network (based on the member's topical interests), where 2000 candidates are retrieved from a pool of hundreds of millions candidate with a latency budget of a few milliseconds and inbound QPS of several thousand per second. This paper presents a novel retrieval approach that fine-tunes a large causal language model (Meta's LLaMA 3) as a dual encoder to generate high quality embeddings for both users (members) and content (items), using only textual input. We describe the end to end pipeline, including prompt design for embedding generation, techniques for fine-tuning at LinkedIn's scale, and infrastructure for low latency, cost effective online serving. We share our findings on how quantizing numerical features in the prompt enables the information to get properly encoded in the embedding, facilitating greater alignment between the retrieval and ranking layer. The system was evaluated using offline metrics and an online A/B te

Semantic search with large language models (LLMs) enables retrieval by meaning rather than keyword overlap, but scaling it requires major inference efficiency advances. We present LinkedIn's LLM-based semantic search framework for AI Job Search and AI People Search, combining an LLM relevance judge, embedding-based retrieval, and a compact Small Language Model trained via multi-teacher distillation to jointly optimize relevance and engagement. A prefill-oriented inference architecture co-designed with model pruning, context compression, and text-embedding hybrid interactions boosts ranking throughput by over 75x under a fixed latency constraint while preserving near-teacher-level NDCG, enabling one of the first production LLM-based ranking systems with efficiency comparable to traditional approaches and delivering significant gains in quality and user engagement.

Large Language Model (LLM) agents are increasingly used in real-world products, where personalized and context-aware user interactions are essential. A central enabler of such capabilities is the agent's long-term semantic memory system, which extracts implicit and explicit signals from noisy longitudinal behavioral data, stores them in a structured form, and supports low-latency retrieval. Building industrial-grade long-term memory for LLM agents raises five challenges: scalability, low-latency retrieval, privacy constraints, adaptability, and observability. We introduce the Hierarchical Long-Term Semantic Memory (HLTM) framework, which organizes textual data into a schema-aligned memory tree that captures semantic knowledge at multiple levels of granularity, enabling scalable ingestion, privacy-aware storage, low-latency retrieval, and transparent provenance; HLTM further incorporates an adaptation mechanism to generalize across diverse use cases. Extensive evaluations on LinkedIn's Hiring Assistant show that HLTM improves answer correctness by more than 5% and retrieval F1 by more than 10%, while significantly advancing the Pareto frontier between query and indexing latency. HLT

Generative artificial intelligence (GenAI) systems rely heavily on user-generated data for training. As governments and platforms impose increasing restrictions on the use of personal data, an important question is whether limiting access to user data for AI training affects the performance of AI-enabled economic systems. We examine this question in the context of labor-market matching. Our setting exploits a unique sequence of LinkedIn policy changes: the quiet introduction of user data collection for AI training in August 2024, the restriction of Hong Kong user data from AI training in October 2024, and the subsequent restoration of data access in November 2025. Using employment and job-posting data from Revelio Labs and a Difference-in-Differences design comparing Hong Kong and Singapore, we find that the restriction significantly increased labor-market frictions: employee turnover increased and tenure declined, vacancies remained open longer, job-posting match rates fell, and wages decreased. To strengthen the causal interpretation, we examine the full policy lifecycle: labor market outcomes improved following LinkedIn's initial AI rollout, deteriorated after the restriction, a

In this paper, we present a novel method for detecting fake and Large Language Model (LLM)-generated profiles in the LinkedIn Online Social Network immediately upon registration and before establishing connections. Early fake profile identification is crucial to maintaining the platform's integrity since it prevents imposters from acquiring the private and sensitive information of legitimate users and from gaining an opportunity to increase their credibility for future phishing and scamming activities. This work uses textual information provided in LinkedIn profiles and introduces the Section and Subsection Tag Embedding (SSTE) method to enhance the discriminative characteristics of these data for distinguishing between legitimate profiles and those created by imposters manually or by using an LLM. Additionally, the dearth of a large publicly available LinkedIn dataset motivated us to collect 3600 LinkedIn profiles for our research. We will release our dataset publicly for research purposes. This is, to the best of our knowledge, the first large publicly available LinkedIn dataset for fake LinkedIn account detection. Within our paradigm, we assess static and contextualized word emb

Data Driven Attribution, which assigns conversion credits to marketing interactions based on causal patterns learned from data, is the foundation of modern marketing intelligence and vital to any marketing business and advertising platform. In this paper, we introduce a unified transformer-based attribution approach that can handle member-level data, aggregate-level data, and integration of external macro factors. We detail the large scale implementation of the approach at LinkedIn, showcasing significant impact. We also share learnings and insights which are broadly applicable to the marketing and ad tech fields.

A post embedding (representation of text in embedding space that effectively captures semantic meaning) is a foundational component of LinkedIn that is consumed by product surfaces in retrieval and ranking (e.g., ranking posts in the feed or video tab). This paper presents the post embeddings used at LinkedIn, where a pre-trained transformer-based large language model (LLM) is taken as input and fine-tuned using multi-task learning across a diverse set of semantic labeling tasks. We observe positive transfer, leading to improved performance across all tasks, compared to training them independently. The generated post embeddings outperform baseline models in zero-shot learning, demonstrating its potential for broader applicability. Furthermore, the generated post embeddings' performance surpasses that of OpenAI's ADA-001 and ADA-002 embeddings on LinkedIn specific datasets and tasks. We also describe the offline evaluation methodology and the deployment to our near-line infrastructure, which makes the post embedding available for use within minutes of post creation for any downstream application. We present how the embeddings were applied in the Feed product surface, in both ranking

Recent advancements in learning-based query performance prediction models have demonstrated remarkable efficacy. However, these models are predominantly validated using synthetic datasets focused on cardinality or latency estimations. This paper explores the application of these models to LinkedIn's complex real-world OLAP queries executed on Trino, addressing four primary research questions: (1) How do these models perform on real-world industrial data with limited information? (2) Can these models generalize to new tasks, such as CPU time prediction and classification? (3) What additional information available from the query plan could be utilized by these models to enhance their performance? (4) What are the theoretical performance limits of these models given the available data? To address these questions, we evaluate several models-including TLSTM, TCNN, QueryFormer, and XGBoost, against the industrial query workload at LinkedIn, and extend our analysis to CPU time regression and classification tasks. We also propose a multi-task learning approach to incorporate underutilized operator-level metrics that could enhance model understanding. Additionally, we empirically analyze th

We conduct an independent, third-party audit for bias of LinkedIn's Talent Search ranking system, focusing on potential ranking bias across two attributes: gender and race. To do so, we first construct a dataset of rankings produced by the system, collecting extensive Talent Search results across a diverse set of occupational queries. We then develop a robust labeling pipeline that infers the two demographic attributes of interest for the returned users. To evaluate potential biases in the collected dataset of real-world rankings, we utilize two exposure disparity metrics: deviation from group proportions and MinSkew. Our analysis reveals an under-representation of minority groups in early ranks across many queries. We further examine potential causes of this disparity, and discuss why they may be difficult or, in some cases, impossible to fully eliminate among the early ranks of queries. Beyond static metrics, we also investigate the concept of subgroup fairness over time, highlighting temporal disparities in exposure and retention, which are often more difficult to audit for in practice. In employer recruiting platforms such as LinkedIn Talent Search, the persistence of a particu

Large Language Models (LLMs) have made it easier to create realistic fake profiles on platforms like LinkedIn. This poses a significant risk for text-based fake profile detectors. In this study, we evaluate the robustness of existing detectors against LLM-generated profiles. While highly effective in detecting manually created fake profiles (False Accept Rate: 6-7%), the existing detectors fail to identify GPT-generated profiles (False Accept Rate: 42-52%). We propose GPT-assisted adversarial training as a countermeasure, restoring the False Accept Rate to between 1-7% without impacting the False Reject Rates (0.5-2%). Ablation studies revealed that detectors trained on combined numerical and textual embeddings exhibit the highest robustness, followed by those using numerical-only embeddings, and lastly those using textual-only embeddings. Complementary analysis on the ability of prompt-based GPT-4Turbo and human evaluators affirms the need for robust automated detectors such as the one proposed in this study.

In this paper, we present a QUBO formulation designed to solve a series of generalisations of the LinkedIn queens game, a version of the N-queens problem, for the Takuzu game (or Binairo), for the most recent LinkedIn game, Tango, and for its generalizations. We adapt this formulation for several particular cases of the problem, as Tents \& Trees, by trying to optimise the number of variables and interactions, improving the possibility of applying it on quantum hardware by means of Quantum Annealing or the Quantum Approximated Optimization Algorithm (QAOA). We also present two new types of problems, the Coloured Chess Piece Problem and the Max Chess Pieces Problem, with their corresponding QUBO formulations.

In this paper, we present LiGNN, a deployed large-scale Graph Neural Networks (GNNs) Framework. We share our insight on developing and deployment of GNNs at large scale at LinkedIn. We present a set of algorithmic improvements to the quality of GNN representation learning including temporal graph architectures with long term losses, effective cold start solutions via graph densification, ID embeddings and multi-hop neighbor sampling. We explain how we built and sped up by 7x our large-scale training on LinkedIn graphs with adaptive sampling of neighbors, grouping and slicing of training data batches, specialized shared-memory queue and local gradient optimization. We summarize our deployment lessons and learnings gathered from A/B test experiments. The techniques presented in this work have contributed to an approximate relative improvements of 1% of Job application hearing back rate, 2% Ads CTR lift, 0.5% of Feed engaged daily active users, 0.2% session lift and 0.1% weekly active user lift from people recommendation. We believe that this work can provide practical solutions and insights for engineers who are interested in applying Graph neural networks at large scale.

The literature has shown that combining a few non-Personal Identifiable Information (non-PII) is enough to make a user unique in a dataset including millions of users. This work demonstrates that a combination of a few non-PII items can be activated to nanotarget users. We demonstrate that the combination of the location and {5} rare ({13} random) skills in a LinkedIn profile is enough to become unique in a user base of {$\sim$970M} users with a probability of 75\%. The novelty is that these attributes are publicly accessible to anyone registered on LinkedIn and can be activated through advertising campaigns. We ran an experiment configuring ad campaigns using the location and skills of three of the paper's authors, demonstrating how all the ads using $\geq13$ skills were delivered exclusively to the targeted user. We reported this vulnerability to LinkedIn, which initially ignored the problem, but fixed it as of November 2023.%This nanotargeting may expose LinkedIn users to privacy and security risks such as malvertising or manipulation.