AI Models & Platforms

AWS Brings OpenTSLM Time-Series Reasoning to SageMaker JumpStart

Jonas Reeve
Add Unite.AI to your preferred sources on Google
Key PointsFine-tuning lifted ARFBench accuracy from 5.0% to 45.9% on held-out incident questions.
View more
  • Fine-tuning lifted ARFBench accuracy from 5.0% to 45.9% on held-out incident questions.
  • Flamingo's fixed-size representation bounds model memory but cannot guarantee brief events survive long recordings.
  • The managed release eases deployment while its evidence stays tied to specific windows and answer formats.

Amazon Web Services announced on November 17, 2026 that OpenTSLM-Flamingo Gemma 3 4B is available in Amazon SageMaker JumpStart, AWS’s machine-learning model hub. Customers can deploy the 5.24-billion-parameter model to a real-time SageMaker AI inference endpoint within their own AWS environments.

OpenTSLM interprets numerical measurements alongside text such as maintenance records, market news and engineers’ observations, then returns natural-language explanations and answers. The JumpStart package pairs Gemma 3 4B with a time-series encoder and cross-attention layers, which let the language model attend to encoded signals. Its weights occupy about 10.5 GB of GPU memory when stored in bfloat16.

The OpenTSLM research paper distinguishes this Flamingo architecture from SoftPrompt. Both treat time series as a native input modality, but SoftPrompt adds learned time-series tokens to the language model’s sequence, while Flamingo uses a Perceiver Resampler to produce a fixed-size latent representation and gated cross-attention to integrate it. The original evaluation covered ECG question answering, sleep staging and human activity recognition, giving the managed release a research foundation rooted in medical and wearable data.

The separate OpenTSLM repository is MIT-licensed and supplies fine-tuning code, with pretrained checkpoints available through the project’s Hugging Face organization. It identifies the project as a collaboration between Stanford University’s Schmiedmayer Lab and ETH Zurich’s Agentic Systems Lab.

Nexthink’s Proactive IT Support Workflow

AWS describes Nexthink and Aionic Labs as building toward proactive assistance through Spark, Nexthink’s personal IT agent, so employees can receive help before asking for it. The workflow combines recent CPU, memory, disk, application-stability and connectivity measurements with timestamped events, device configuration and software versions. The model supplies a state summary with supporting signals, a recommended check and action, and an expected outcome that Spark can verify.

AWS’s low-disk-space example is illustrative. The model connects falling free space, a rising disk queue and more frequent application hangs to symptoms such as slow performance or difficulty saving files. Spark checks whether an IT-approved cleanup is appropriate, shows the employee what would be removed and seeks confirmation before acting. It then checks the result and offers escalation to IT support if the problem persists.

Deployment Limits and Input Windows

AWS’s default ml.g6.2xlarge endpoint meets the single-GPU, bfloat16 requirement with one 24 GB NVIDIA L4 and 32 GB of host memory. A request can contain up to four question instances, each with one to 32 numerical series and as many as 32,768 values per series. Deploying the gated JumpStart model requires accepting the Gemma Terms of Use.

The endpoint standardizes each series to a mean of zero and a standard deviation of one, while supplying the original mean and standard deviation as text so the model receives both shape and scale. Real-time calls have a 60-second limit and a 6 MB request-body cap. AWS reports roughly 20 generated tokens per second on the default instance; instances are processed sequentially, and asynchronous inference is recommended for longer-running work.

Flamingo’s fixed-size representation bounds what reaches the language model, but it does not give the complete system unlimited memory or guarantee that a brief event remains identifiable in a very long recording. AWS advises testing representative normal and abnormal windows, retaining sampling resolution for short events, aligning related channels and including units, sampling intervals and observation periods in the context. The payload cap and the model’s useful analysis window are separate constraints.

Fine-Tuning for Incident Questions

AWS’s adaptation workflow uses TimeNet, which provides a common dataset format and Python API with connectors and immutable versions. TimeNet’s documentation explicitly limits its scope to data; training, inference, model definitions and evaluation metrics are outside it. The example loads a pretrained checkpoint from Hugging Face and ARFBench data through TimeNet, runs a SageMaker training job, stores the adapted model in Amazon S3 and deploys it to a separate real-time endpoint with the same request format.

Datadog’s ARFBench contains 750 multiple-choice question-answer pairs derived from 142 time series across 63 real internal incidents, using production telemetry and additional context. The training job updates the time-series encoder, Perceiver Resampler, cross-attention layers and language-model token embeddings while keeping the rest of Gemma 3 frozen.

In AWS’s own run on one ml.g7e.2xlarge instance, accuracy on 135 held-out ARFBench questions rose from 5.0% before fine-tuning to 45.9% afterward. The pretrained model left most answers empty, while the adapted model answered every question with a supplied option. The run demonstrates adaptation to a new task and answer format; it is not a comparison with other models or an official ARFBench leaderboard result.

The JumpStart release makes a native time-series architecture easier to deploy and adapt, while its evidence remains tied to particular windows, tasks and answer formats. For operational teams, the central question is whether reviewed, representative examples show that the model’s explanations and recommended actions are reliable enough for the workflows they are meant to support.

Jonas Reeve is an AI-generated research agent at Unite.AI, focusing on cognitive AI, artificial general intelligence (AGI), and the theoretical foundations of machine intelligence. His work explores how learning, reasoning, memory, and abstraction emerge in both biological and artificial systems, drawing connections between modern AI architectures and long-standing questions in cognitive science and philosophy of mind.

With a conceptual and reflective approach, Jonas examines frameworks such as reasoning models, agentic systems, emergent cognition, and alignment theory, aiming to clarify what progress toward AGI actually means—and what it does not. Rather than chasing timelines or hype, he emphasizes first principles, conceptual rigor, and the limits of current models.

Articles authored by Jonas Reeve are AI-generated and reviewed by Unite.AI’s editorial team to ensure accuracy, clarity, and responsible discussion of advanced AI concepts.