About AutoML Services

From uploaded data to results people can understand and review.

Xalec AutoML Services is the flagship product experience from Xalec AI. It turns raw data into understandable, reviewable results through guided uploads, data checks, model comparison, prediction explanations, reporting, and prediction in one connected application.

20 guided analyses Test results before model claims Saved results and source-based reports
Product proof
20
Active service workflows
Cross-sectional, longitudinal, time series, media, survival, ranking, recommendation, and more.
1
One guided path
Upload, validate, train, explain, report, and predict from one product shell.
268
Available model setups
Models users can choose before training; availability may depend on installed software.
What the app is solving

Too many analytics teams still bounce between notebooks, scripts, dashboards, and slides. AutoML Services brings the workflow together so modeling outputs can be reviewed, trusted, and acted on responsibly.

Why it exists

The workflow is usually harder than the model.

Many teams can get a model to run. The harder part is choosing the right workflow, validating the data shape, tracking progress, understanding cost, producing explainable outputs, and turning technical results into something reviewable by stakeholders. AutoML Services is designed around that full workflow rather than only the training step.

Fragmented analytics workflows waste time and trust.
Teams still hand off work between scripts, notebooks, dashboards, and slide decks.
Harder workflows are often underserved.
Longitudinal, repeated-measures, survival, count, and report-heavy use cases need more than generic tabular AutoML.
Stakeholder-ready output does not happen automatically.
Metrics alone are not enough; the app needs explainability, validation, saved outputs, and report generation built in.
One Product, One Guided Path

How users move from raw data to decision-ready outputs.

The app is designed to keep your work connected instead of making you restart on each page.

1. Choose

Guided service selection

Start by data shape and objective instead of forcing every problem into the same upload form.

2. Review

Upload and readiness review

Preview the required columns, role assignment, warnings, and recommended models before spending credits.

3. Run

Curated multi-model evaluation

Train candidate models, compare them on the right metrics, and keep progress visible in the workspace and job page.

4. Extend

Explainability, reports, and prediction

Use saved job outputs for SHAP, AI reporting, exports, and prediction.

What the product emphasizes

Evaluation before hype.

Readiness and quality gates
Users see whether the data and role setup are ready before training begins.
Cost-aware runs
The app shows billing and credit cues near the actions that actually drive cost.
Grounded explainability
Explainability is treated as model-behavior evidence, not as proof of causality.
Reports that stay with the job
Reports, references, figures, and saved notes stay with the job after you leave the page.

Built for harder workflows

The strongest differentiator in the product today is that more complex structured workflows are treated as first-class: longitudinal and repeated-measures analysis, survival, count modeling, ranking, recommendation, and report-heavy work.

One consistent shell

The homepage, services catalog, upload pages, workspace, job detail view, reports, prediction flows, and billing cues are meant to feel like one guided product rather than disconnected tools.

Trust through saved evidence

Saved outputs, validation notes, references, figures, and matching exports matter because trustworthy analytics should be reviewable after the job completes, not only while the page is open.

Who It Is For

Teams that need workflow ROI, not just model novelty.

AutoML Services is useful for analysts, researchers, public health teams, clinical and outcomes researchers, evaluation teams, and technical operators who need a workflow that is easier to use than a custom ML stack but more transparent than a one-click black box.

It is especially strong when the work is messy, governed, report-heavy, or structurally more complex than a standard tabular prediction problem.

What Users Should Expect
Clarity
The app should make the next step visible, especially during upload, job monitoring, and add-on actions.
Confidence
Status, readiness, warnings, and cost cues should be shown before and during each run.
Durability
Saved outputs should back the report, workspace, and downloadable files rather than leaving the browser view as the only truth.
Start Here

See the product in action from the route that fits your goal.

Explore the services catalog if you are starting a new analysis, or open the workspace if you want to review saved jobs and outputs.