Key facts
| Access control | Scoped API keys with rotation; enterprise SSO and RBAC options |
| Deal separation | Separate keys and retrieval indexes per brokerage or deal team |
| Deployment | Cloud, VPC, on-prem or air-gapped for transaction data |
| Auditability | Request, model and response logging for transaction review trails |
| Data grounding | Embeddings and RAG are live for listings, disclosures and contracts |
| Structured output | JSON mode returns listing and disclosure records in fixed schemas |
| Models | 30+ models behind one OpenAI-compatible API |
| Endpoint roadmap | Files, batch and fine-tuning are coming soon |
TL;DR
- Treat transaction files as the highest-sensitivity class after identity data.
- Scope keys per brokerage, deal team and integration.
- Keep client financials and identity documents out of general prompts.
- Log requests, sources and reviewers for anything sent to a client.
- Keep contract, disclosure and wire-related decisions with humans.
How it works, step by step
- Inventory AI use cases across listing copy, market research, contracts and client updates.
- Classify data by deal sensitivity, identity content and contractual restriction.
- Choose deployment per class: cloud, VPC, on-prem or air-gapped.
- Issue per-brokerage and per-deal-team keys with rotation and central inventory.
- Define log fields and retention: request ID, model, sources, output, reviewing broker.
- Approve a model allow-list and ground answers in current listings and disclosures.
- Require human verification for contract terms and payment instructions.
Try it yourself
Open the data residency checker →
Listings, clients and transaction data
Real estate content spans public listings, market data, client preferences, transaction documents, identity records and payment instructions. Each class needs its own rule, and payment instructions deserve special handling because fraud attempts target exactly those messages.
Start with listing copy and market research, then extend to client and transaction files only on deployments that keep prompts, documents and embeddings inside the brokerage or an agreed boundary.
Keys, agents and least privilege
Issue a distinct API key per brokerage, deal team, application and environment. Store keys in a secrets manager, rotate on a schedule, and revoke when an agent leaves or a partner integration ends, which is frequent in a commission-driven industry. Enterprise SSO and RBAC options keep console access aligned with licence status and team changes.
Never place client financial details or identity numbers in prompts where retrieval can supply only the fields a task requires.
Residency, retention and transaction records
Decide where processing happens and how long records live. Region selection covers many needs; VPC, on-prem and air-gapped deployment covers brokerages and markets with stricter requirements. Retention applies to prompts, outputs, logs and retrieval indexes, and transaction records follow their own schedules set by contract and local rules.
Log enough to reconstruct a client-facing output: request ID, model and version, retrieved source identifiers, output and the reviewing broker. See AI audit logs for a schema.
Model governance and brokerage review
Keep an approved model list with evaluation evidence and re-test when versions change. Ground answers in current listings and disclosure documents with citations so agents can verify quickly on a live deal. Contract terms, disclosures and all payment instructions stay with named humans, and the checklist should document verification steps for anything that could move money or create an obligation.
Honest comparison
| Control area | Plugsky capability | Common gap | Owner |
|---|---|---|---|
| Deal separation | Separate keys and indexes per brokerage or deal team | One workspace for all clients | Broker operations |
| Identity | Scoped keys, rotation, enterprise SSO and RBAC options | Shared office credentials | IT security |
| Data boundary | Cloud, VPC, on-prem or air-gapped deployment | Transaction data path undocumented | Compliance |
| Retention | Configurable logging of prompts and outputs | No defined retention | Records |
| Audit trail | Request, model and reviewer logging | Client outputs not traceable | Managing broker |
| Verification | Citations and structured output for agents | AI-altered payment instructions | Finance and legal |
Frequently asked questions
Does using Plugsky make us compliant?
No. Compliance is your program. Plugsky provides deployable controls - scoped keys, deal separation, private deployment, logging - that you document and audit against your own brokerage and privacy obligations.
How do we handle wire fraud risk?
Never let AI generate or alter payment instructions. Keep verification with named humans, require out-of-band confirmation, and treat any AI-drafted message containing account details as untrusted.
What should we log?
Request IDs, model names and versions, retrieved sources, outputs and the reviewing broker, retained under your policy so client-facing work can be reconstructed.
Can client financial data stay private?
Yes. VPC, on-prem and air-gapped deployment keeps prompts, documents and embeddings inside your environment, and the API stays OpenAI-compatible.
Can we use AI for listing descriptions?
Yes, grounded in current listing facts and disclosure requirements with human review before publication.
Is fine-tuning available for our templates?
Fine-tuning, files and batch endpoints are coming soon. Today, use retrieval over approved templates with JSON mode for consistent output structure.
Where should a pilot start?
Pilot on public listing data and market research with the free plan, validate quality and logging, then extend to client and transaction workflows on private deployment.