Author

ByteDance SE Lab

Official account of ByteDance SE Lab, sharing research and practical experience in software engineering. Our lab unites researchers and engineers from various domains to accelerate the fusion of software engineering and AI, driving technological progress in every phase of software development.

Articles

Likes

121

Views

Comments

Latest from ByteDance SE Lab

45 recent articles

ByteDance SE Lab

Jul 16, 2026 · Cloud Computing

From Data Lake to State Lake: Redesigning Storage Infrastructure for the Agent Era

The article analyzes how the rise of AI agents transforms storage needs from massive, static data lakes to fast, state‑centric lakes, outlining three workload phases, new storage paradigms, and the three pillars—Sandbox Store, Artifact Store, and Agent Observation & Evaluation—that enable scalable, low‑latency, and observable agent execution.

AI AgentsArtifact StoreCloud Storage

0 likes · 22 min read

From Data Lake to State Lake: Redesigning Storage Infrastructure for the Agent Era

ByteDance SE Lab

Jul 14, 2026 · Databases

Instant DB Creation & Auto‑Scale: How Volcano PostgreSQL Serverless Supercharges Feishu MiaoDa AI Development

The article explains how AI‑generated code can be delivered in seconds, but traditional databases add a ten‑minute bottleneck, and shows that Volcano Cloud PostgreSQL Serverless reduces instance creation to 10 seconds, speeds up branch operations 200‑fold, offers scale‑to‑zero, cuts costs by 99.99%, and provides a Data‑as‑Git workflow with a step‑by‑step recovery demo.

AI DevelopmentCloud DatabaseData as Git

0 likes · 12 min read

Instant DB Creation & Auto‑Scale: How Volcano PostgreSQL Serverless Supercharges Feishu MiaoDa AI Development

ByteDance SE Lab

Jul 13, 2026 · Artificial Intelligence

How Viking AI Search Uses Personalization to Boost Conversion

Viking AI Search enhances user‑centric results by first ensuring keyword relevance, then adjusting recall and ranking with item popularity scores and personalized interest tags, offering configurable weak or strong interventions that let developers quickly add effective, controllable personalization without building complex models.

E-commerceViking AIitem popularity

0 likes · 8 min read

How Viking AI Search Uses Personalization to Boost Conversion

ByteDance SE Lab

Jul 13, 2026 · Artificial Intelligence

Turning Agents into an Audio‑Video Workbench with AI MediaKit CLI + Skill

The AI MediaKit CLI and Skill, announced at the 2026 Force conference, expose over 100 audio‑video capabilities as a unified, agent‑friendly workbench, enabling natural‑language driven editing, enhancement, and delivery of videos through structured commands, long‑task handling, and edge‑cloud collaboration.

AI MediaKitAgentAudio Video Processing

0 likes · 11 min read

Turning Agents into an Audio‑Video Workbench with AI MediaKit CLI + Skill

ByteDance SE Lab

Jul 10, 2026 · Artificial Intelligence

How a Non‑English Seller Used a Bot and SearchCLI to Conquer Japan, Korea, and English‑Speaking Markets

Faced with language barriers in cross‑border e‑commerce, a seller with no English, Japanese, or Korean skills built a multilingual storefront by combining the no‑code web builder Kouzi with Viking AI SearchCLI, automating data ingestion, multilingual search, recommendation, QA, evaluation, and one‑click deployment to sell across Japan, Korea and English‑speaking regions.

Data IngestionE‑commerceMultilingual AI

0 likes · 11 min read

How a Non‑English Seller Used a Bot and SearchCLI to Conquer Japan, Korea, and English‑Speaking Markets

ByteDance SE Lab

Jul 9, 2026 · Information Security

How Claude Mythos Reshapes Enterprise Security Architecture

Claude Mythos, Anthropic's 2026 agency‑grade model, pushes autonomous vulnerability discovery and exploitation from months to minutes, forcing enterprises to accelerate operations, tighten zero‑trust boundaries, and redesign security architectures for resilience against AI‑driven attacks.

AI red teamAI securityAgent-based attacks

0 likes · 24 min read

How Claude Mythos Reshapes Enterprise Security Architecture

ByteDance SE Lab

Jul 8, 2026 · Databases

Volcano Milvus Hits Nearly 3× VectorDBBench Leader in Retrieval Speed

Under a fixed monthly budget of about 7,100 CNY, Volcano Milvus combines DiskANN and RaBitQ (including Extended‑RaBitQ) with in‑memory layout and query‑path slimming to deliver 20,420 QPS, 2.5 ms average latency, 93.9 % recall, achieving nearly three times the VectorDBBench top score while using roughly one‑third of the memory, disk and compute resources of competing solutions.

DiskANNMilvusRaBitQ

0 likes · 12 min read

Volcano Milvus Hits Nearly 3× VectorDBBench Leader in Retrieval Speed

ByteDance SE Lab

Jun 17, 2026 · Information Security

Server Firmware Security Practices for AI-Infra: Threat Modeling, Trusted Boot, and Large‑Scale Remediation

The article analyzes the rising firmware security challenges of AI‑Infra servers, presents a full‑machine threat model, outlines trusted‑boot and measurement architectures, shares a large‑scale CVE‑2023‑34335 remediation case, and discusses tools and long‑term security evolution for heterogeneous server fleets.

AI infrastructureBoardSentinelSecure Boot

0 likes · 24 min read

Server Firmware Security Practices for AI-Infra: Threat Modeling, Trusted Boot, and Large‑Scale Remediation

ByteDance SE Lab

Jun 4, 2026 · Artificial Intelligence

How ContextBucket Gives Agents Unlimited Memory and a Unified Workspace

The article analyzes the context fragmentation challenges faced by production‑grade agents—memory loss, workspace inconsistency, and governance gaps—and explains how ContextBucket’s managed service unifies memory and workspace with hybrid retrieval, serverless storage, and multi‑tenant isolation, delivering significant accuracy and cost improvements.

Agent MemoryHybrid RetrievalLLM Performance

0 likes · 15 min read

How ContextBucket Gives Agents Unlimited Memory and a Unified Workspace

ByteDance SE Lab

May 19, 2026 · Artificial Intelligence

Introducing Uni-Agent: veRL’s Open‑Source Unified Framework for General‑Purpose Agent Training

Uni-Agent is an open‑source framework that unifies building, running, and training of general AI agents, offering extensible model, tool, and environment modules, scalable sandbox execution via veFaaS, live monitoring, and demonstrated performance gains on large‑scale coding‑agent experiments.

AgentScalable ExecutionUnified Framework

0 likes · 8 min read

Introducing Uni-Agent: veRL’s Open‑Source Unified Framework for General‑Purpose Agent Training