Question 1

What is Delfa?

Accepted Answer

Delfa builds AI agents that automate patient recruitment for clinical trials, the single operational bottleneck behind much of trial cost and delay: with 22,000+ drug candidates in the pipeline but only about 5,000 new trials a year, more than a third of the cost comes down to finding, qualifying, and enrolling patients, still a manual funnel of Excel trackers, cold calls, and phone-collected histories.

Delfa replaces that human bottleneck with agents trained on clinical trial documentation and guardrailed accordingly, which speak with patients in any language by phone, SMS, or chat to prescreen for medical eligibility, answer nuanced inclusion and exclusion questions, collect full medical histories (calls regularly run 20 minutes), book visits, and write directly into a site's CRM or CTMS. Quality is enforced heavily: every patient interaction is logged, reviewed, and scored across 20+ QA metrics, physicians audit outputs, and alerts fire in real time when something looks off.

Because patients have opted in to trial contact, the use case is comparatively clean. Delfa already powers 50+ live trials with top site networks and has deployed production AI for J&J, Medtronic, and Exscientia. Reported outcomes include a migraine study site exceeding its enrollment target by 25% while saving 20 recruiter hours at an 11% lead-to-enrolled rate, and a vaccine trial where the agent matched human conversion. The company plans to build an AI-native CRM for patient recruitment on top of the voice wedge. $3.8M from Air Street.

Question 2

How much does Delfa cost?

Accepted Answer

Enterprise sales to clinical research sites and site networks; no rates published in retrieved materials.

Question 3

What does Delfa integrate with?

Accepted Answer

AI agents write directly into a site's CRM or CTMS after each interaction and support native recruitment-progress monitoring from HubSpot; in one documented pattern, Delfa's voice agent completes a pre-screening call and passes the transcript via API to a partner browser agent that logs into the clinical management system and fills out the full screening questionnaire, with exceptions flagged for human review.

Question 4

Who is Delfa for?

Accepted Answer

clinical research sites, site networks (Flourish, Elevate, Re:Cognition), pharmaceutical sponsors, CROs

Question 5

What are the best alternatives to Delfa?

Accepted Answer

Based on documented capability coverage across the Agentic Index 14 point taxonomy, the closest alternatives to Delfa among voice agents are Consio, careCycle, Liberate, Casca. Each is graded on the same public evidence, and no vendor pays for placement.

Question 6

Does Delfa give you tracing and observability into agent runs?

Accepted Answer

Full under the Agentic Index verification standard. Delfa publishes public evidence of tracing, observability and audit logging into agent runs, so a team can see what an agent did and why. Evidence basis: Every patient interaction is logged, reviewed, and scored across 20+ QA metrics with physician audits, real-time alerts, and full audit trails, plus native recruitment monitoring from HubSpot, Air Street and Asteroid materials 2026-07-22. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Question 7

Does Delfa require human approval before an agent acts?

Accepted Answer

Full under the Agentic Index verification standard. Delfa publishes public evidence of human approval steps and guardrails, so an agent action can be held for review before it executes. Evidence basis: Every interaction is logged, reviewed, and scored across 20+ QA metrics, physicians audit outputs, real-time alerts fire when something looks off, and exceptions are flagged for human review, Air Street and Asteroid materials 2026-07-22. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Question 8

Does Delfa support MCP and tool calling?

Accepted Answer

Partial under the Agentic Index verification standard. Delfa publishes some public evidence of tool calling and API or SDK extensibility, but narrower than full Model Context Protocol (MCP) support as defined by the Agentic Index taxonomy. Evidence basis: Delfa passes call transcripts to downstream agents via API in a documented integration, indicating an API surface, though no public SDK or MCP is documented, Asteroid materials 2026-07-22. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Integrations & Tool Calling	Full
Agents write directly into a site's CRM or CTMS and support native recruitment monitoring from HubSpot, with API handoff of call transcripts to downstream systems, Air Street and Delfa materials 2026-07-22
Workflow Orchestration	Full
Agents fulfill the full recruiter role end to end: prescreen for eligibility, answer inclusion/exclusion questions, collect full medical histories, book visits, and write into the CRM/CTMS, Air Street and Delfa materials 2026-07-22
Knowledge Grounding & RAG	Full
Agents are trained on clinical trial documentation and speak about inclusions and exclusions with nuance, grounding screening in each trial's protocol, Air Street materials 2026-07-22
Human Oversight & Guardrails	Full
Every interaction is logged, reviewed, and scored across 20+ QA metrics, physicians audit outputs, real-time alerts fire when something looks off, and exceptions are flagged for human review, Air Street and Asteroid materials 2026-07-22
Security, Identity & Governance	Partial
Documented HIPAA-grade reliability with full audit trails in the integration workflow, short of named attestations in retrieved materials, Asteroid and Delfa materials 2026-07-22
Observability & Auditability	Full
Every patient interaction is logged, reviewed, and scored across 20+ QA metrics with physician audits, real-time alerts, and full audit trails, plus native recruitment monitoring from HubSpot, Air Street and Asteroid materials 2026-07-22
Memory & State Persistence	Partial
Agents retain patient histories and progress to prescreen, schedule, and remind the same patients across the recruitment funnel, implying retained per-patient state short of a documented memory layer, Delfa materials 2026-07-22
Deployment & Data Residency	Unable to verify
Cloud platform; no self hosting or data residency options documented, Delfa materials 2026-07-22
Prebuilt Agents, Templates & Packs	Partial
Agents are configured per trial protocol from clinical documentation, a productized recruiter role adapted per study, short of a self-serve template library, Air Street and Delfa materials 2026-07-22
Triggers & Channel Coverage	Full
Agents reach patients in any language by phone, SMS, or chat for outbound outreach and reminders as well as inbound engagement, 24/7, Air Street and Delfa materials 2026-07-22
Model Flexibility & Routing	Unable to verify
No customer facing model choice or routing documented, Delfa materials 2026-07-22
APIs, SDKs & MCP Extensibility	Partial
Delfa passes call transcripts to downstream agents via API in a documented integration, indicating an API surface, though no public SDK or MCP is documented, Asteroid materials 2026-07-22
Testing, Debugging & Optimization	Partial
The 20+ QA metric scoring with physician audit functions as a continuous quality-evaluation loop over agent outputs, short of a documented customer-facing testing or simulation surface, Air Street materials 2026-07-22
Browser & Computer Use	Unable to verify
Delfa itself handles the voice conversation; the browser-based CTMS data entry in the documented workflow is performed by a separate partner agent, not Delfa, Asteroid materials 2026-07-22

Delfa

Vendor details

Canonical URL

Category

Funding status

Company status

Use cases & customers

Primary use cases

Target customers

Deployment options

Integrations

Sources & related URLs

Research sources

Agentic Index coverage score

Pricing

Related vendors

Alternatives to Delfa

Head to head

Contact us