All Services

Normalises string fields and computes an equal-width histogram for quoted_amount or quoted_units across up to 200 procurement records, returning two analytical reports.

Applies string normalisation and trailing mean series analysis to up to 200 procurement research records for quoted_amount or quoted_units fields, returning two analytical reports.

Computes a trailing moving average series and adjacent row-to-row changes for procurement records (quoted_amount or quoted_units) in a single API call returning two analytical reports.

Computes a trailing moving average time series and a date-by-category pivot table for up to 200 procurement records (quoted_amount or quoted_units), returning two analytical reports in one call.

Computes a trailing moving-average time series and an equal-width frequency histogram for procurement quoted_amount or quoted_units data supplied by the caller.

Computes a trailing moving-average time series and fills missing values in procurement records (quoted_amount or quoted_units) for up to 200 rows, returning two analytical reports.

Computes trailing mean time series and keyed dataset comparison reports for procurement records containing quoted amounts and units, returning two analytical reports for up to 200 rows.

Computes a trailing mean time series and a secondary-measure weighted mean for up to 200 procurement records (quoted_amount, quoted_units) using local analytics — no external lookups.

Runs two local analytical reports on procurement records: a Pearson/rank correlation between quoted_amount and quoted_units, and a date-by-category pivot table, for up to 200 rows.

Computes a secondary-measure weighted mean over procurement records (up to 200 rows), weighting one numeric field by another locally without external enrichment.

Normalises supplier names, dates, and numeric fields in procurement records for up to 200 rows, returning cleaned and standardised data.

Searches the current web for alternative products, feature comparisons, and migration options tailored for platform operator use cases.

Searches the current web for product alternatives, feature comparisons, and migration guides tailored to research team workflows.

Searches the current web for product alternatives, competitor tools, and feature comparison resources specifically tailored for software developers and builders.

Searches live public web pages for information about product catalog management tailored to agent builders, returning source links and snippets.

Searches live public web pages about product catalog management topics, returning relevant source links and snippets tailored for analyst audiences.

Searches live public web pages about product catalog management topics specifically focused on technology buyers, returning source links and text snippets.

Computes confidence calibration bins for a product category classifier (apparel, home, electronics, other) given up to 200 labelled prediction records with class probabilities.

Computes a confusion matrix for a product category classifier (apparel, home, electronics, other) from up to 200 labeled text records with actual and predicted labels.

Generates a fixed-holdout learning curve for product category text classification (apparel, home, electronics, other) using a naive-Bayes baseline on up to 200 labelled records.

Audits up to 200 labelled text records for a product category classifier (apparel, home, electronics, other), computing evaluation metrics and naive-Bayes baselines.

Computes label distribution statistics for a supplied set of labelled product-category text records (apparel, home, electronics, other)

Computes macro, micro, and weighted aggregate evaluation scores for a product category classifier (apparel, home, electronics, other) given up to 200 labelled text records with actual and predicted labels.

Computes the multiclass Brier score for product category text classification predictions across apparel, home, electronics, and other classes.

Computes per-label precision and recall metrics for a product category classifier across up to 200 labelled text records with actual and predicted labels.

Computes prediction error records (e.g. misclassification details, per-record loss) for a naive-Bayes product category text classifier across apparel, home, electronics, and other categories.

Evaluates a supplied set of labelled product category predictions (apparel/home/electronics/other) broken down by user-defined groups, returning per-group classification metrics.

Computes top-k label coverage metric for a product category classifier across up to 200 labelled text records with 4 classes (apparel, home, electronics, other).

Trains a naive-Bayes text classifier on up to 200 labelled product-category records and classifies new query texts into apparel, home, electronics, or other.

Searches the live web to find supported versions, compatibility matrices, and integration information for procurement software products.

Searches the live web for supported version ranges, compatibility matrices, and integration support information relevant to product management contexts.

Searches the live web to find supported versions, compatibility matrices, and integration information for products, tailored for research teams.

Computes adjacent row-to-row changes for product feature telemetry data (events or unique users) across up to 200 typed records, returning a local descriptive analytics report.

Computes a base-100 index series from product feature telemetry rows, normalizing events or unique-user counts relative to a chosen baseline row for local descriptive analytics.

Computes category-level frequency counts across product feature telemetry records (up to 200 rows) and returns a local descriptive analytics report.

Computes a daily sum time series for product feature telemetry data (events or unique users) across up to 200 records, grouped by feature and date.

Pivots up to 200 product feature telemetry records by date and category, producing a local descriptive analytics report of events or unique users per feature over time.

Identifies and reports duplicate row keys in a batch of product feature telemetry records (up to 200 rows).

Identifies and removes exact duplicate rows from product feature telemetry data (up to 200 records) and returns a bounded deduplication report.

Fills missing numeric values (events or unique_users) in product feature telemetry records using zero, mean, or median imputation, returning a bounded local analytics report.

Computes a grouped descriptive statistics summary (e.g. min, max, mean, count) for product feature telemetry records, segmented by feature group.

Compares two keyed datasets of product feature telemetry records (up to 200 rows each) and returns local descriptive analytics on differences between them.

Computes a descriptive numeric distribution report (histogram, percentiles, summary stats) over up to 200 product feature telemetry records for a chosen metric field.

Filters product feature telemetry rows to only those whose selected numeric metric (events or unique_users) falls within a specified min–max interval, returning up to 200 matching records.

Computes a secondary-measure weighted mean (e.g. events weighted by unique_users) across product feature telemetry records, returning a local descriptive analytics report.

Computes a trailing (rolling) mean series over time-series product feature telemetry records, returning a smoothed time-series for events or unique users per feature group.

Computes a two-measure correlation (events vs. unique_users) across up to 200 product feature telemetry records, returning a local descriptive analytics report with no enrichment or forecasting.

Validates a product import CSV for blank/empty rows and returns a blank row report, supporting up to 1000 rows with configurable delimiter and category allowlists.

Validates calendar date columns in a product imports CSV file, checking up to 1000 rows against declared date constraints and category rules.

Validates that all cells in a product import CSV do not exceed a specified maximum character length, returning a structured audit report.

Validates a declared CSV table contract for product import data, checking column structure, categories, and value ranges for up to 1000 rows.

Validates a product CSV against declared schema rules and returns a report of all invalid rows, including which rows fail and why.

Validates a product import CSV file and returns a located error report pinpointing cell-level data issues, up to 1000 rows.

Validates that numeric columns in a product import CSV contain properly typed numeric values, flagging non-numeric cells and range violations.

Validates that record keys in a product import CSV are unique, checking up to 1000 rows for duplicate keys against a declared table contract.

Validates that required cells in a product-import CSV are present and non-empty, checking up to 1000 rows against a declared schema contract.

Validates a product import CSV for spreadsheet formula injection risks, flagging cells that contain formula-like patterns across up to 1000 rows.

Validates and normalises a product import CSV (up to 1000 rows), trimming whitespace and coercing cell types, returning a typed conversion report.

Validates a product import CSV file (up to 1000 rows) and returns a structured summary report covering data quality, format compliance, and category membership.

Validates a product import CSV for whitespace issues in up to 1000 rows, checking headers, cell lengths, categories, and numeric ranges.

Searches live DuckDuckGo results to find current public pricing pages and plan details for consulting-related products and services.

Searches the web via DuckDuckGo to find current public pricing pages and plan details for software/SaaS products, tailored for product team use cases.

Searches live DuckDuckGo results to find current public pricing pages and plan details for software products, focused on research team use cases.

Detects and reports duplicate paragraphs in a product specification document supplied as Markdown or plain text.

Masks email addresses and phone numbers found in product specification documents (Markdown or plain text) using deterministic pattern-based redaction.

Converts Markdown or plain text product specification documents (up to 20,000 characters) into an escaped HTML preview using deterministic document templates.

Audits the heading hierarchy of a product specification document (Markdown or plain text) to identify structural issues in heading organization.

Generates a structured heading outline from a supplied product specification document in Markdown or plain text.

Extracts and inventories all HTTP links from a product specification document (Markdown or plain text, up to 20,000 characters) using deterministic processing.

Analyzes a product specification document (Markdown or plain text) to produce a deterministic length and paragraph profile report.

Scans a product specification document and produces an inventory of all quoted passages found within it.

Reorders recognised sections in a product specification document (Markdown or plain text) into a canonical sequence using deterministic template-based processing.

Audits a product specification document to verify all required sections are present and properly populated

Extracts a specific named section from a product specification document provided as Markdown or plain text

Generates a structured section template for a product specification document from supplied Markdown or plain text input.

Normalises whitespace in Markdown or plain-text product specification documents, removing inconsistencies and producing a clean, standardised output.

Counts and ranks word frequencies in a product specification document supplied as Markdown or plain text

Validates a JSON string and returns whether it is valid, along with its parsed representation

Formats and pretty-prints raw JSON text into human-readable indented output

Computes a SHA-256 cryptographic hash digest of a given input text string

Encodes a plain text string into its Base64 representation

Extracts action-marker lines (e.g. TODO items) from a project brief document supplied as Markdown or plain text

Splits a project brief document (Markdown or plain text) into bounded text chunks of a specified maximum character size using deterministic processing.

Detects and reports duplicate paragraphs within a project brief document supplied as Markdown or plain text.

Masks email addresses and phone numbers found in a project brief document using deterministic pattern matching, returning the sanitized text.

Converts a project brief written in Markdown or plain text (up to 20,000 characters) into an escaped HTML preview, safe for embedding or display.

Audits the heading hierarchy of a project brief document (Markdown or plain text) to ensure proper structure and nesting.

Extracts and returns a structured heading outline from a supplied project brief document in Markdown or plain text.

Extracts and inventories all HTTP links found in a project brief document (Markdown or plain text, up to 20,000 characters) using deterministic processing.

Analyzes a project brief document (Markdown or plain text) to produce a length and paragraph profile, including structural metrics like word count, character count, and paragraph breakdown.

Generates a deterministic paragraph-level change report comparing a new project brief against a previous version, highlighting additions, removals, and modifications.

Reorders the recognised sections of a project brief document into a standard sequence using deterministic template-based processing.

Audits a project brief document to verify all required sections are present and properly structured

Extracts and returns discrete sections from a project brief document supplied as Markdown or plain text

Applies a deterministic section template to supplied Markdown or plain text for a project brief document

Normalises whitespace in project brief documents (Markdown or plain text) using deterministic processing without generative AI.

Computes word frequencies for a project brief document supplied as Markdown or plain text, up to 20,000 characters.

Adds or subtracts a specified number of whole days from an ISO date, tailored for project workflow scheduling.

Allocates a specified meeting duration across a list of supplied topics, generating a structured agenda tailored for project workflows.

Distributes a list of project tasks across a specified number of available days, generating a day-by-day project plan.