Video summary

DataSift Open API Official Release / Integrate Any Tool You Want

Main summary

Key takeaways

Technology

Summary

DataSift announced the release of its Open API, positioning the platform as a human-facing CRM with a headless backend that AI agents, scripts, and custom applications can access. The API works with the same account data as the regular interface and covers both CRM operations and nationwide property data.

API capabilities and access

  • The API has 303 documented endpoints: 272 for CRM and prospect management, and 31 for property-data lookups.
  • Supported operations include managing records, owners, phone numbers, tags, lists, statuses, custom fields, filter presets, tasks, boards, notes, messages, call history, and activity logs. The API also supports property and sales-history lookups.
  • API keys inherit the permissions of the account user who creates them and can be revoked. The presenters cautioned that API access can make consequential changes, including deleting account data, so keys and automated actions should be managed carefully.
  • The API can be used without AI through code or scripts. AI tools can also help users generate scripts and interact with the API.

Custom-tool demonstration

Tyler demonstrated a locally run prospecting and calling tool built with Cursor, DataSift’s API, and Twilio. It allowed the user to:

  • Load records using saved filter presets.
  • View property details, notes, and phone-number tags.
  • Call through records, record dispositions and notes, skip numbers, and advance to the next number.
  • Update DataSift records in real time, including changing statuses and logging activity.
  • Create separate work queues so team members can work on different records without overlapping.

According to Tyler, the prototype took about four hours to build. He said the queue and focus-mode concept had been added to the product roadmap.

AI agents and workflow examples

The presenters distinguished an AI assistant or script from an AI agent. In their description, an agent combines an AI model, a playbook or instructions, tools, scheduled checks or triggers, and scripts that carry out tasks. Agents can respond to events or run on a schedule, while people handle conversations and judgment-heavy work.

Ty described four agents operating through Slack:

  • Atlas analyzes and retrieves data, including first-to-market records, auction information, tax delinquency updates, and high-priority properties.
  • Diggs checks new records, performs skip tracing and data cleanup, and helps identify records needing deeper prospecting.
  • Roma handles marketing-related workflows and routes positive responses to the human lead manager.
  • Quinn provides KPI reporting and supports other operational tasks.

Examples shown included agents coordinating foreclosure and probate prospecting, routing records to calling or marketing workflows, reporting new records and team KPIs in Slack, and stopping outreach when a seller responds positively. The presenters emphasized that agents can hand work to one another and that users can start with a single agent or a few specific skills rather than reproducing the full setup.

Guidance and resources

The webinar included a basic first-connection tutorial:

  1. Create an API key in DataSift’s settings under Integrations → Open API.
  2. Store the key in an environment-variable file rather than sharing it directly in prompts.
  3. Provide an AI tool with the API documentation at developers.datasift.ai.
  4. Run a simple “hello world” or smoke test, such as asking how many records are in the account.

Tyler demonstrated the process with Claude and reported that the test connected successfully and returned an account record count. He recommended caution about giving an AI unrestricted access and advised structuring workflows rather than relying on unreviewed actions.

The speakers also presented a Real Estate AI Operator Boot Camp, a course with 11 modules and about 200 videos, alongside five days of live work focused on building four agents. They described a separate in-person event in Orlando and said a guide or PDF covering agent workflows and use cases would be shared. Attendees were encouraged to read the guide and share tools they build in the community.

Product philosophy and business guidance

The speakers advocated using AI for operational work—such as data cleanup, skip tracing, underwriting preparation, KPI reporting, and workflow coordination—while keeping people involved in seller conversations, trust-building, negotiation, and closing. They argued that users should learn to build and adapt their own tools rather than depend entirely on a vendor’s roadmap or a closed, bundled AI agent.

The webinar also included market commentary. Tyler cited investor-purchase and off-market transaction figures to argue that opportunities remain. These figures were presented by the speaker and were not independently assessed during the webinar.

Reviews

Tyler noted that DataSift has Trustpilot reviews and asked users to leave reviews. The webinar did not discuss specific review ratings or individual customer feedback.

Main speakers and sources

  • Tyler Austin — DataSift CEO/CPO; presented the API, setup tutorial, custom-tool demo, and product philosophy.
  • Ty Garrett — demonstrated the Slack-based AI-agent workflows used in his real estate business.
  • Attendees — briefly contributed ideas and questions in the webinar chat.

Rate this summary

Your feedback will help improve summaries.

Improve this summary

Reprocess with a stronger model when the summary feels incomplete or inaccurate.

Pro

Translate summary in another language

Pro

Ask questions to this video

Chat for follow-up questions, clarifications, and source-backed answers.

Coming soon

Share this summary

Original video