Latest News
Joint Research on Precipitation Estimation to Be Presented at AGU26
Joint research on applying an AI model for precipitation estimation to radar-sparse regions will be presented as a poster at AGU26.
Neurogica Launches General API Availability of yoft, a Time-Series Foundation Model for Operational Data
We have developed yoft, a time-series foundation model for operational data such as electricity demand, sales and equipment sensor readings, and opened general API access for forecasting, anomaly detection and imputation at yoft.ai. On the public GIFT-Eval benchmark, yoft ranks first among methods that use no additional training.
Neurogica Will Exhibit at CEATEC 2026 (October 13–16, Makuhari Messe)
Neurogica will exhibit at CEATEC 2026, held at Makuhari Messe from October 13 to 16. At our booth (Hall 4, 4H425) we will introduce forecasting, anomaly detection and imputation with yoft, our time-series foundation model.
New tech blog post: Testing Our Service for Vulnerabilities with Chinese Open-Weight Models
Amid a wave of data breaches, we are reviewing our unreleased service with publicly available Chinese open-weight models and Altar-1, a GLM derivative. Models accessible to attackers can also support defenders. We discuss why we run these models ourselves and report two runs per model: Qwen and GLM each produced one verified finding. We examine remediation priorities, false positives, and incomplete runs.
New tech blog post: Slam Dunk Was Read More on Announcement Day Than on Release Day: Four Shapes of Booms in Wikipedia Pageviews
On the day THE FIRST SLAM DUNK opened, the Japanese Wikipedia article on Slam Dunk was read 7.9 times as often as usual. On the day the film was announced, it was 21.8 times. Measuring the shape of attention across 36 items, from obituaries and anime to buzzwords and Heisei-era revivals, yields four types. The arrival of a peak cannot be read from the views before it, but the decay after a peak can be roughly read from its first seven days.
New tech blog post: The First Step Toward Time Series Forecasting: OLTP and OLAP
Running aggregation queries against the production database makes them compete with ordinary requests for CPU and disk. How do you pick an analytics database, and how do you get data into it? This post lays out four sync methods and four stores, then benchmarks the same query over 11.68M rows across five setups.
New tech blog post: Forecasting Power Demand with Prophet and Time Series Foundation Models: What Mattered Was Temperature, Not the Model
"Isn't Prophet enough for demand forecasting?" Some might think so. On nine years of daily peak demand in TEPCO's service area, Prophet did not significantly beat simply copying the value from seven days earlier. What mattered was not the choice of model but how temperature was fed in.
New tech blog post: Why Ordinary Models Cannot Forecast Irregularly Sampled Data
"If the sampling is irregular, just resample to a regular grid before feeding the model." That is half right: ordinary models cannot take irregular data as is, so without resampling the accuracy does not come out. But resampling measures accuracy on data from which the timing of each observation has been thrown away, and that distortion never shows up in the accuracy number. This article checks both claims with synthetic data and PhysioNet.
New tech blog post: The Evolution of Irregular Time Series Models: From GRU-D and Neural ODE to mTAN and Beyond
"Can't you just fill in the gaps before feeding the model?" Some might think so. From GRU-D and Neural ODE to mTAN and ISTS-PLM, research on irregularly sampled time series has shifted from how to impute missing values to how to represent irregular time inside the model. This article traces that shift.
New tech blog post: Forecasting JEPX Prices with Time Series Foundation Models: The Baseline to Beat Was Yesterday (Part 1)
"Isn't a classical statistical model enough for electricity price forecasting?" European studies suggest so. A year of zero-shot evaluation across ten JEPX spot price series shows the baseline to beat is simply copying yesterday, and only the time series foundation models beat it. Importing the standard European benchmark design leads to the wrong conclusion.
New tech blog post: What Is a 1% Gain in Forecast Accuracy Actually Worth?
"We improved MAE by 5%." And how much money is that worth? The link between forecast accuracy and value has been studied since 1951. Minimizing an error metric and optimizing a decision are not the same problem.
New tech blog post: Can Large Language Models Forecast Time Series?
"Can't today's AI already do time series forecasting?" The push to forecast time series with LLMs began in 2023 and is still unsettled in 2026. Can they, or can't they? A time series researcher separates what the papers actually showed from personal opinion.
Neurogica Passes the Screening Round of NEDO Challenge, Baggage-Loading Robot: Contest 1 (Baggage Identification)
Neurogica has passed the screening round of Contest 1 (Baggage Identification) in NEDO's prize-based program for airport ground handling, as one of seven companies selected to proceed.
We have renewed our corporate website
We have completely redesigned our corporate website to better convey our technology and ongoing initiatives.
Two papers on time-series forecasting and vital sign reconstruction accepted to ICASSP 2026
Papers from our team on time-series forecasting and PPG vital sign reconstruction were accepted to ICASSP 2026, a top conference in signal processing.
We have relocated our headquarters
Following business expansion, we have relocated our headquarters to 7F Gloria Shibuya Building, 2-19-17 Shibuya, Shibuya-ku, Tokyo.