Published June 30, 2026

June 2026 product update

June 30, 2026

In June, we launched three new API categories, Analytics, Datastore and Clubs, shipped external secrets manager support for credential storage, expanded event delivery across accounting and recruiting integrations, and added 52 new integrations.

Analytics and Datastore extend Unified into product analytics and database access: two categories that put real-time data, fetched directly from the source, in front of AI agents, dashboards, and reporting features without per-vendor SDKs or custom integration logic.

Three new categories in a single month, bringing Unified to 500+ integrations across 30 categories, is the pattern that separates next-generation unified APIs from first-generation ones: real-time, pass-through architecture means breadth across categories no longer comes at the cost of depth within them.

Analytics API

We launched a new Analytics API category for reading and writing traffic, events, sessions, visitors, and performance reports across analytics platforms through one category-scoped API.

Supported integrations: Google Analytics, Mixpanel, PostHog, Pendo, YouTube Analytics, and Visier.

The API provides consistent objects across these integrations:

Object	Purpose
Property	The site, app, or analytics account being measured
Event	A tracked action: page view, click, form submit, purchase, sign-up, or video play
Session	A visit, with duration, landing and exit pages, source, and conversion flags
Visitor	An identified or anonymous user, with engagement totals and first and last seen timestamps
Report	Aggregated metrics over a date range, broken down by dimension
Method coverage:

Google Analytics: list, get, remove for Property; list for Event; list for Report
Mixpanel: list, get, create for Event; list, get, create, update for Visitor; list, get for Property; list for Report; created and updated [virtual webhooks](/blog/unlock_real_time_data_with_virtual_webhooks) for Visitor
PostHog: list, get, create, update, remove for Property; list for Event; list, get for Session; list, get, create, update for Visitor; list for Report; list, get, update, remove for Employee
Pendo: list, get for Property; list, get, create for Event; list, get, create, update for Visitor; list, get for Report
YouTube Analytics: list, get for Property; list for Report
Visier: available through raw pass-through

Reporting is metric-driven. You request a KPI, scope it to a property and date range, and break it down by dimension, through one endpoint: GET /analytics/{connection_id}/report. The type selector covers traffic (USERS, SESSIONS, PAGE_VIEWS), engagement (BOUNCE_RATE, ENGAGEMENT_RATE), conversion and value (CONVERSIONS, REVENUE, TRANSACTIONS), and video (VIDEO_VIEWS, WATCH_TIME) metrics; the dimension selector breaks results down by DATE, SOURCE, COUNTRY, DEVICE_TYPE, CAMPAIGN, and more. The same metric and dimension vocabulary works across every integration above, so dashboard code is written once and runs regardless of which platform a customer connects.

Each request is executed directly against the source API: no caching, no sync jobs, no stored records. Traffic is regionalized across US, EU, and AU for data-residency requirements. The connection_id changes. The API calls don't.

What this enables:

Read normalized traffic, conversion, and engagement reports across analytics platforms with one metric and dimension vocabulary
Drill into visitor-level and event-level detail where the platform exposes it
Ingest events through a normalized event schema with UTM and geo fields
Feed real-time traffic and conversion signals to AI features and copilots

This was the most-requested new category. It is net-new: teams previously built and maintained a separate integration per analytics platform.

→ Unified Analytics API: Read Traffic, Events, and Reports Across Platforms

→ How to Build a Cross-Platform Analytics Dashboard: Google Analytics, Mixpanel, PostHog

Datastore API

We launched a new Datastore API category for reading, querying, and writing structured data across spreadsheets, operational databases, and cloud data warehouses through one category-scoped API.

Supported integrations: Airtable, Google BigQuery, Google Sheets, MongoDB, MongoDB Data API, Notion, Snowflake, and Supabase. Airtable, Google BigQuery, MongoDB Data API, and Supabase are new this month; Snowflake, Notion, Google Sheets, and MongoDB gained Datastore object coverage.

The API provides consistent objects across these integrations:

Object	Purpose
Database	The workbook, project, schema, or data store connection
Table	A table, sheet, view, or collection, with field definitions and types
Record	A single row or document, with typed field values
Query	A structured query or raw SQL executed against the source
Method coverage:

Snowflake: list, get, create, remove for Database; list, get for Table; list, get, create, update, remove for Record; create for Query; created and updated virtual webhooks for Record
Google BigQuery: Database, Table, Record, and Query objects
Notion: list, get for Database; list, get for Table; list, get, create, update, remove for Record; create for Query
Google Sheets: list, get for Database; list, get for Table; list, get, create, update for Record; create for Query
Airtable: list, get for Database; list, create, update, remove for Table; list, get, create, update, remove for Record
MongoDB: list, get, create, remove for Database; list, get, create, remove for Table
Supabase: Database, Table, Record, and Query objects; OAuth2 sign-in authentication

The Query object is what unifies a spreadsheet and a data warehouse under one interface. It runs in two modes: structured filters that are portable across every integration (conditions like eq, gt, lt, like, in, and contains, combined with AND/OR, plus sorting, pagination, and aggregation through count, sum, avg, min, max, and group_by), and raw SQL via query.sql where the platform supports it, for joins, window functions, and warehouse-specific work. One endpoint, POST /datastore/{connection_id}/query, runs both. Records come back with typed field values (string, number, boolean, date, or selection), so you work with structured data rather than raw JSON.

This opens Unified to data-engineering and ETL use cases: large-scale querying, extraction, and record-level reads and writes against warehouses and structured stores. The same Datastore endpoints work across every integration above. The connection_id changes. The API calls don't.

All requests are stateless and routed directly to the source API: no caching, no sync jobs, no stored records. Traffic is regionalized across US, EU, and AU for data-residency requirements.

What this enables:

Read and write records across Snowflake, BigQuery, MongoDB, Supabase, Airtable, Notion, and Google Sheets through one Datastore API
Run one query model across every platform: structured filters for portable logic, raw SQL where the warehouse supports it
Build AI agents that read, filter, and write structured data back across the platforms a customer already uses
Receive change events on records (Snowflake virtual webhooks)

→ Unified Datastore API: Read, Query, and Write Structured Data Across Spreadsheets, Databases, and Warehouses

→ How to Build AI Agents on Customer Warehouse Data: Snowflake, BigQuery, and Supabase

Clubs API

We also launched a Clubs API for reading and writing sports club data. New integrations this month: Strava and TeamSnap, joining Rev Sport.

The API provides unified Group, Member, Activity, Event, and Location objects, with the same real-time, pass-through reads and writes as every other category, including created and updated virtual webhooks for members (Rev Sport).

Secrets Managers

Integration credentials are the most sensitive data in any integration: OAuth client secrets, access and refresh tokens, API keys, and private keys. Teams with strict security or compliance requirements often need these stored in their own infrastructure rather than a vendor's database.

Unified now stores connection and workspace credentials in your own external secrets manager. Unified keeps only the reference needed to locate the secret and read or refresh it at call time. The credential value itself stays in your vault. Non-sensitive metadata, such as integration type and environment name, stays in Unified.

Supported secrets managers:

AWS Secrets Manager
Azure Key Vault
Google Cloud Secret Manager
HashiCorp Vault

Composio is also supported as a read-only credentials store for end-customers who have already authorized accounts there and do not want to re-authorize. Composio continues to manage its own OAuth2 token refresh.

You can create more than one secrets manager per workspace and control which one is used:

Set a workspace default for new credentials
Override the default per environment, such as Production or Sandbox
Assign a specific secrets manager to a specific authorized connection

When Unified stores or refreshes a credential, it resolves the target in order: a secrets manager assigned directly to the credential, then the environment's configured secrets manager, then the workspace default. If none applies, Unified stores the credential encrypted in its own database. Changing the configuration moves affected credentials in the background, and existing connections keep working during the move.

What this enables:

Store production and sandbox credentials in separate vaults from a single workspace
Give each of your customers a dedicated environment that routes to that customer's own vault, so credentials are never shared across customers
Meet region-specific or environment-specific credential-residency requirements at the infrastructure level

→ Using Secrets Managers in Unified

New integrations

52 integrations added in June across analytics, data stores, recruiting, HR, advertising, and more.

Analytics

Datastore

ATS / Recruiting

HR & Directory

Advertising

CRM / Call Center / Ticketing

Knowledge Management

Mintlify

Generative AI

Alibaba Cloud Model Studio

Tasks

Condens

Shipping

Trexity

E-Commerce / Accounting / Payments

Lightspeed Restaurant (K-Series)

Clubs

Enhancements

Event delivery expansion

We expanded native and virtual webhook coverage across more integrations, moving more of Unified toward detect → deliver event patterns.

Accounting saw the largest expansion:

Microsoft Dynamics 365 Business Central: created, updated, and deleted native webhooks across Employee, Invoice, Bill, Salesorder, Purchaseorder, Account, Contact, Journal, and Transaction; created and updated virtual webhooks across Employee, Company, Bill, Organization, Order, Salesorder, Purchaseorder, Account, Contact, Journal, and Transaction
MYOB: created and updated virtual webhooks across Employee, Order, Invoice, Journal, Account, Transaction, Contact, Salesorder, and Purchaseorder

Additional created and updated event coverage:

NetSuite: Contact
Snowflake: Record
Mixpanel: Visitor
Freshteam: Group, Timeoff, Candidate, Application
Welcome to the Jungle: Candidate, Company
Slack: Recording
QuickBooks: Inventory
Pinpoint: Document
Microsoft Teams: Recording
Eploy: Job
TrackerRMS: Document
EasyPost: Tracking
Rev Sport: Member
BambooHR: created, updated, and deleted events for Employee
Twilio: created and updated native webhooks for Call; created native webhooks for Recording

Picklist object

We added a Picklist object across CRM and ATS integrations, giving you read access to the valid picklist and dropdown values defined in the source system. Added with list method coverage to: Salesforce, HubSpot, Pipedrive, NetSuite, Microsoft Dynamics 365 Sales, Zoho CRM, Zoho Recruit, Bullhorn, Loxo, Lightfield, Gainsight, AgencyZoom, and Rev Sport.

This removes a common source of failed writes: you can read the accepted values for a field before writing to it, rather than hard-coding enums per vendor.

New objects and method coverage

QuickBooks: Inventory object (list, get, update) with created and updated virtual webhooks; total_stock readable field on Item
Deputy: Location, Group, Timeshift, and Timeoff objects with full list, get, create, update, remove coverage
Stripe: create and update methods for Item, with writable fields including type, prices, name, metadata, duration, and collections
Google Search Ads 360: Organization, Campaign, Group, and Ad objects with list and get coverage
FactorialHR: Course and Activity objects; create, update, remove for Activity
Personio, Freshteam, Certn, JobAdder: new object coverage across Company, Group, Timeoff, Job, Candidate, Application, and Location

Field-level expansions

ATS original_substatus****: added as a readable field on Application across 11 integrations, including Greenhouse, Greenhouse (Ingest), Ashby, Rippling, Workday (Legacy), UKG Ready, Trakstar Hire, JobScore, CareerPlug, Workstream, and Atlas
Workday: manager_id and employment_status readable on Employee; type, duration_type, and duration readable on Timeoff; termination_reason and terminated_at writable on Employee
Marketing company****: added as a readable and writable field on Member across Klaviyo, Yotpo, Instantly.ai, and ActiveCampaign
Jobvite: metadata readable and writable across Job and Candidate
Sage Intacct: size and data readable on File
FactorialHR: timeoff_days_used, timeoff_days_total, pronouns, locations, and groups readable on Employee

Pagination improvements

Broad pagination improvements shipped across integrations this month. Notable additions:

Microsoft Dynamics 365 Business Central: updated_gte, org-scoped, and virtual webhook pagination variants across Employee, Company, Bill, Order, Salesorder, Purchaseorder, Account, Contact, Journal, and Transaction
MYOB: updated_gte and virtual webhook pagination across all accounting objects
Microsoft Dynamics 365 Customer Engagement: limit pagination across 13 objects
Microsoft Teams: full pagination suite for Recording, including call, user, and time-range scoping
Google Ads: query pagination for Campaign, Ad, and Creative
BambooHR: limit, offset, and slow_fields for Timeoff

Final takeaway

June brought three new categories, Analytics, Datastore and Clubs, into Unified, expanded native and virtual webhook coverage across accounting and recruiting integrations, and added a Picklist object that removes a recurring source of failed writes.

The direction is consistent: more integrations behaving predictably, more objects covered completely, and less custom logic required per vendor. Two new categories in a single month is what a real-time, fully unified architecture produces when breadth no longer costs depth: the part first-generation unified APIs can't follow.

→ Start your 30-day free trial

→ Book a demo

All articles