Integrating Certificial's API Into Your Platform:
Frequently Asked Questions

Yes. We create and assign sandbox credentials for your dev team, giving you access to an isolated testing environment where you can hit our open APIs and mock the integration. We also provide sample API responses in our documentation so you can see the data structure before sandbox access. Let us know when you're ready to build and we'll provision the environment.

Each of your customers will use their own production API credentials (not shared partner credentials), so the sandbox is purely for connector development and testing.

The API returns consistent, structured data regardless of whether the agency is in our partner network. You receive the certificate (COI), any requested endorsement documents, all certificate data fields, and compliance information at three levels: high-level status (compliant/non-compliant), policy-level compliance for each coverage type, and data-element-level compliance showing exactly which fields meet or don't meet requirements.

When an agency isn't networked and manually uploads a PDF, we use AI to extract the data and return it in the same structured format, so you don't need to build different processes for networked vs. non-networked responses.

You can integrate gradually. We only request and track insurance for the vendors you explicitly tell us to handle, so you control the rollout pace. Many partners start with a pilot group, test the workflow, and expand from there. You can run your existing COI process and Certificial in parallel, directing specific vendors or clients to Certificial when ready. No "big bang" cutover required.

We enhance, not replace. Out of 25 integrations completed, none required full replacement. Certificial handles the parts that are manual and repetitive (uploading COIs, entering coverage limits, tracking renewals) automatically through the agent network. What remains in your module are usually client-facing features like dashboards, reporting, and workflow management, now displaying richer, real-time data from Certificial.

The best way to determine the exact approach is to show us a demo of your current COI module so we can map which features Certificial would handle vs. which stay in your system.

Yes, activity logging via API is currently in development. It will provide details on every communication: who was emailed, what triggered it, date/time, and status.

Until then, some partners use a workaround: our outreach follows a predictable schedule (initial request, 7-day follow-up to the insured if no agent response, etc.), so you can hard-code standard timestamps based on when the request was initiated. The API will automatically override those placeholders with actual data once live. We can also provide documentation on our outreach cadence so you can set proper customer expectations.

We follow a structured escalation process: Day 0 (initial agent request), Day 7 (if no response, we start emailing the insured to confirm their agent info is correct), then ongoing monitoring and escalation. Our team tracks non-responsive cases to identify issues like wrong contacts or agencies that no longer represent the vendor.

We typically consider a request "stalled" after 2-3 weeks of no response, but you control how your platform handles those cases, whether that's flagging for manual review, allowing a PDF upload, or marking the vendor non-compliant pending documentation.

Pass a policy request ID to our "cancel tracking" API endpoint. It immediately cancels the request, removes data from our UI and active systems, and returns a success confirmation for your audit trail.

Note: because the insurance industry is heavily regulated, we cannot permanently delete all records. Data is tagged as "canceled/deleted" in our database but retained for regulatory compliance, as agencies may need to prove they issued a certificate to a specific party years later. This is standard industry practice. From your customers' perspective, the data is no longer accessible or active once you hit that endpoint.

80-85% of all insurance agencies are in our network, including the top 100 brokers who issue ~90% of all commercial insurance. That's 26,000-27,000 out of 28,000 agencies writing commercial policies in the U.S.

For agencies not yet networked, we still process their certificates: they upload a COI through our system, we extract the data using AI, and you receive the same structured information via API. The difference is that networked agencies provide automatic, real-time monitoring (policy changes, cancellations, renewals flow directly from their system), while non-networked agencies require us to request updated certificates manually. We've never had an agency refuse to issue a certificate through our platform.

~60% comes from exclusive integration with Applied Systems (formerly Applied Epic), the largest AMS provider. ~30% comes from one additional system where we have access but not yet exclusive integration (may change soon). The remaining ~10% is distributed across 5-6 other AMS platforms, all with exclusive access. In total, ~70% of our coverage is exclusive to Certificial.

When you send a request through our API, the agent receives it in the system they already use every day. We don't ask them to log into a separate portal or change their workflow, which is why adoption is high.

95-96% for networked agencies when the data provided upfront is clean (correct agent email, accurate vendor info). We've maintained this rate with partners processing 1,500+ COI verifications per week. We've never had an agency outright refuse to use our platform.

The few cases that don't connect immediately are typically data quality issues (wrong agent contact, agency no longer represents that vendor), which our team monitors and resolves. For context, our network has been tested at scale with clients like Amazon, FedEx, and UPS.

Yes to both. We reach out to non-networked agencies directly. They upload the certificate through our platform, we extract the data using AI, and return it in the same structured format as networked responses. Simultaneously, our agency sales team contacts them about joining the network (free for most agencies, streamlines their workflow, reduces repetitive requests).

If they join after your initial request, that vendor automatically upgrades to real-time monitoring going forward with no action needed on your end.

We work with the top 100 brokers issuing ~90% of commercial insurance. We can't confirm specific broker names publicly due to NDAs, but our network is comprehensive at the enterprise level. We integrate at the AMS level, not broker by broker, so if a major broker uses Applied Systems, Vertafore, or another AMS we're integrated with, their agents are automatically in our network.

We don't accept certificates directly from vendors/carriers. Certificates must come from the licensed agent or broker, which is how we ensure legitimate, verified coverage data.

For networked agencies (80-85% of cases) with clean data: 80% of requests fulfilled in under 72 hours, many in 24-48 hours. Once the agent clicks to share the policy data, it flows to you immediately. For non-networked agencies: typically 5-7 business days, since we need to reach out, request the certificate, and wait for a response.

The biggest factor in turnaround is data quality. Accurate vendor info, correct agent email, and complete details upfront lead to fast results. Messy data (wrong agent, outdated contacts) means our team has to investigate and correct, which adds time.

Networked: 24-72 hours. The request appears directly in the agent's AMS and sharing is a button click. Non-networked: 5-7 business days, since it requires email outreach, manual responses, and potential follow-ups.

Even our non-networked process is often faster than traditional methods because we have established agency relationships, follow up proactively, and escalate to the insured after 7 days if the agent hasn't responded. Over time, as more agencies join our network, the percentage of fast responses increases.

We automatically escalate after 7 days by contacting the insured (vendor/carrier) directly to confirm agent info and encourage follow-up. Our team monitors requests in real time and can manually intervene to expedite high-priority cases.

The fastest path: clean data upfront (correct agent email, complete vendor details). With clean data and a networked agency, expect 24-48 hours. For partners with urgent SLA requirements, we recommend flagging high-priority requests so our team can prioritize them.

Networked agencies: real time, often within minutes to hours. We're integrated into the agent's management system through the Smart COI platform, so policy changes (coverage limits, cancellations, renewals) automatically flow through Certificial to you via API. No request needed; the system continuously monitors and pushes updates.

Non-networked agencies: no real-time monitoring, so updates come only when you submit a new request or the agency proactively sends an updated certificate. This is why network coverage matters: it eliminates the gap between when a policy changes and when you know about it. With traditional methods, you might not learn about a cancellation for weeks. With networked agencies, you're notified immediately.

Same as the overall certificate request (24-72 hours networked, 5-7 days non-networked). If the agent shares driver and VIN info as part of the policy data, it flows through immediately with the rest of the structured data in the same API response. No additional delay.

The caveat: we can't force agents to share info they don't have or aren't required to provide. Some policies include named drivers and VINs, some don't. Availability depends on whether that information exists in the policy itself.

Gradually. We only request and track insurance for the specific vendors you tell us to handle. Many partners start with a pilot (one customer, one region, or one product line), test the integration, gather feedback, and expand. You can run your existing COI process and Certificial in parallel for as long as needed. No "big bang" migration required.

Some partners keep both systems running indefinitely for different customer segments (Certificial for real-time monitoring, legacy system for simpler needs).

Yes. Each customer or vendor you add is independent. New API requests get processed without any system-wide change or reconfiguration. Existing tracked vendors continue operating normally. You might start with 10 customers in Month 1, add 50 in Month 2, and scale to hundreds over six months. Each addition is just another API call.

The only consideration is whether your internal processes (onboarding, training, support) can handle the volume as you scale. From a Certificial perspective, there's no limit.

Phase 1 (Pilot), 4-8 weeks: Select 1-3 customers or 50-100 vendors. Validate API data flow, UI display, and customer experience.

Phase 2 (Controlled Expansion), 2-3 months: Expand to 10-20 customers or a specific segment (one region, one product tier). Refine processes, train support, document common questions.

Phase 3 (Broad Rollout): Open to all customers or make it a standard feature. Some partners keep both systems in parallel indefinitely, offering Certificial as a premium feature or opt-in upgrade. You control the pace.

You can start fresh. Just begin sending new vendor requests via API and we handle them going forward. Historical data stays in your existing system for reference and compliance.

If you want real-time monitoring for existing vendors (rather than waiting for renewals), you can upload them in bulk with basic vendor info and agent contact details. This is optional. Some partners do a clean cutover, others bring over their entire active vendor base. Either works, and you can decide customer by customer.

Yes. Many partners offer Certificial as a premium feature for customers who want real-time monitoring and automated compliance, while keeping a legacy COI upload for customers with simpler needs. You could also segment by risk: high-risk vendors go through Certificial, low-risk or one-time vendors use the legacy process.

Certificial operates independently via API, so you decide which requests to send to us and which to handle internally. No technical limitation on running hybrid approaches.

Certificates must come exclusively from licensed insurance agents or brokers. We never accept them directly from vendors or carriers. Fraud rates run as high as 25-30% when certificates come from vendors (they can easily create fakes with AI, Photoshop, etc.), so requiring everything to flow through the agent eliminates the primary fraud vector. Learn more about how to detect fraudulent COIs.

Even when a non-networked agency uploads a PDF, we verify it's from a legitimate, licensed agency before accepting it and cross-reference the extracted data against known agency information.

Four-layer vetting process: (1) License validation through the DOI licensing board for their state. (2) Domain, phone, and physical address verification against public records and insurance databases. (3) Producer (agent) affiliation check to confirm they're actually tied to that agency. (4) Signature and formatting comparison against our proprietary database of known legitimate agency patterns.

This is especially important in freight and logistics, where producer fraud is rising. We've seen fake agencies creating domains nearly identical to real ones (e.g., "MarqueeIG.com" vs. "MarqueeInsuranceGroup.com"). Our vetting catches these before fraudulent certificates reach your system.

The certificate is rejected before it enters your system. We validate the agency through our vetting process (DOI license, domain, producer verification). If validation fails, we flag it and don't accept the certificate.

Our team then investigates: sometimes it's a typo or outdated contact (we follow up with the vendor to correct it), sometimes it's a genuine fraud attempt (we notify you). Either way, you never receive unverified data. This is especially critical in auto hauling and freight, where "fly-by-night" carriers are more common.

Standard process: everything must come from the licensed agent or broker. However, we can enable vendor upload as a configurable fallback (e.g., unresponsive agent, international vendors not in our network) with these caveats: (1) We still extract and attempt to verify the agent information against our database. (2) Vendor-uploaded certificates are flagged differently so you know they haven't gone through standard agent verification. (3) We use the upload to identify the agent and reach out for future renewals and monitoring.

Vendor uploads, if enabled, are marked as lower-trust and we work to upgrade them to agent-verified as quickly as possible.

Industry studies suggest 25-30% of certificates in construction and logistics may be fraudulent or contain false information. We've had clients discover vendors submitted certificates for coverage they didn't have, or from policies canceled shortly after issuance.

The risk: if a vendor causes damage, injury, or loss and their insurance is fake or invalid, your customer may be liable for the full claim, potentially hundreds of thousands or millions of dollars, plus reputational damage and legal liability. Our agent-only verification model addresses this by confirming directly with the licensed agent that coverage is real, active, and matches what the vendor claims.