Humanly

Nine connected systems appear with logos on the homepage: Workday, SAP SuccessFactors, Oracle Taleo, iCIMS, Greenhouse, Avature, UKG, Dayforce and Employ, under the statement that Humanly connects with the customer's existing ATS, job boards and HR systems with no ripping and replacing, and that the integrations are two way.

The integrations act across three classes. Applicant tracking and HRIS: screening data, dispositions and transcripts are written back so the hiring record stays in one place, with Humanly sitting on top of the existing system as an engagement layer. Calendar: the AI Recruiter "syncs with recruiter and hiring manager calendars to book interviews automatically," with no email back and forth, an authenticated write into two parties' calendars. Job boards and distribution: candidate reach extends across job boards and social surfaces. Those are writes into systems Humanly does not own.

Every named partner sits in the recruiting stack, so this is depth in one cluster, and no connector marketplace or integration directory is published.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

A five stage flow runs across distinct parties. As the homepage describes it, Humanly engages every candidate, screens for real fit, conducts structured interviews, schedules the next step, and delivers ranked candidates with transcripts and clear reasoning, as one system from first message to scheduled interview, set against five disconnected tools. Screening and scheduling run through configurable workflows, so the sequence is the customer's to define rather than fixed.

The participants are distinct: the candidate is conversed with, the recruiter receives a ranked shortlist, and the hiring manager's calendar is booked. The AI Recruiter syncs with recruiter and hiring manager calendars, so two internal parties plus the applicant are coordinated in one pass, and the vendor's persona pages separate recruiters, talent leaders, operations leaders and candidates. Work passes between named parties across ordered stages, rather than one agent taking many steps in a single surface.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

Two maintained corpora sit under the agent. The one that matters is the customer's own: a Talent CRM and built in ATS hold the organization's candidates, pipelines, transcripts and dispositions, and the platform "re-engages past applicants" from that store when a new role opens. Rediscovery only works if the record persists between runs and stays queryable. Alongside it sits a network of 600M+ profiles for sourcing on the vendor side.

The agent reads both in every conversation, giving dynamic answers and job details tailored per role, and returning a ranked shortlist with transcripts behind every name. The corpus persists, scales across an entire applicant history far past any context window, and is queried rather than assembled per job.

The grounding is described in outcome terms on the homepage and AI Recruiter page rather than as a documented retrieval structure, and no citation tracking is documented.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

Recruiters get a live exit from the agent, but no gate over what it does. The first item is strong: an "instant handoff to a human chat," so a candidate conversation can leave the agent mid flow and land on a person. Alongside it are recruiter overrides on the agent's output and "configurable knockout criteria," the customer's own boundary on what the agent may advance or reject.

None of that is a review and approve gate over the agent's actions. Humanly's proposition is that the agent engages, screens, scores and books the interview without a recruiter in the loop (a candidate applying at 9pm is on a calendar the next day), and no approval queue, confidence threshold, hold before an invitation is sent, or route to reverse a knockout is documented. Configurable criteria bound the agent in advance; they do not gate it in flight.

Sourcehumanly.io/solutions/screen and homepageread 2026-09-08

Security posture is not published on Humanly's own pages. The site navigation has no security or trust entry, and the footer carries only a Privacy Policy and Terms and Conditions, with no trust center, compliance page or certification badges. The homepage FAQ asks "Is Humanly secure?", but the answer does not display on the page.

Humanly names Microsoft, MGM Resorts, Mutual of Omaha, Trinity Health and Reid Health as customers, and a healthcare and financial services buyer base does not usually buy without an attestation (an inference). A directory listing that says no certifications, SSO or access control matrix are documented is not evidence of absence either.

Sourcehumanly.io homepage navigation, footer and FAQread 2026-09-08

The agent's conversations are recorded, not just summarized. The homepage promises a "ranked shortlist with transcripts behind every name," and ranked candidates delivered "with transcripts and clear reasoning," set against recommendations with no evidence behind them. On a conversational screener the transcript is what the agent did, and structured scoring against a rubric and knockout criteria makes the basis of each decision reconstructable.

No retained action log, routing record, override capture, version history or export is documented on a product page. Timestamps, routing decisions, recruiter overrides and audit export appear only in Humanly's own AI recruiting tools roundup, a marketing article rather than a product page, and are not repeated on any product surface.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

The agent has no documented memory of its own: no memory layer, cross session state store, retained preference set, or memory scope and lifetime is published.

The Talent CRM and built in ATS hold candidates, pipelines, transcripts and dispositions, and past applicants are contacted again from that store, but that is the application's own data: deleting it deletes the hiring record itself.

No retention or deletion control over what the agent has learned, as distinct from candidate data, and no store of taught criteria is published. Configurable knockout criteria are configuration the customer sets in advance and can change, not state the agent accumulates.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

Humanly publishes no deployment choice. No region, regional instance, EU or other residency option, VPC, customer tenant or selection surface appears on the homepage, in the primary navigation or in the footer, which carries only a Privacy Policy and Terms and Conditions. No cloud provider or hosting region is disclosed either.

The product's shape supplies the default answer: it is a hosted conversational service for candidates delivered over SMS, chat and video.

Sourcehumanly.io homepage, navigation and footerread 2026-09-08

Five products each get their own page, but Humanly sells them as one. The primary navigation lists AI Recruiter, AI Interviewer, AI Notetaker, Talent CRM and ATS, each with its own page. Three are agents doing distinct jobs at different points of the funnel, and each stands alone: remove the Notetaker and the Interviewer is still whole.

Beneath them sits a real template layer: an Interview Builder with best practice interview templates a team customizes for questions, depth, tone and scoring. Five solution surfaces (Source, Attract, Screen, Schedule, Interview) and ten industry pages sit alongside as configurations.

The integration is the pitch. The homepage says "everything you need, built into one platform" and "one system, from first message to scheduled interview," set against five disconnected tools, and nothing indicates the three agents can be adopted, declined or swapped separately. No browsable agent catalog, gallery or clone and configure surface is published, and the Interview Builder is a library of interview templates, not of agents.

Sourcehumanly.io/products/ai-interviews and homepage navigationread 2026-09-08

The trigger is the product's central claim, and it is event driven. Humanly greets "every applicant within seconds" on SMS, chat or email, with no waiting and no drop offs, and the homepage states the shape: a candidate applies at 9pm and is on the recruiter's calendar the next day. Nothing waits for a recruiter to start it. A second event path contacts past applicants again from the CRM when a role opens, and a third runs configurable workflows that evaluate, score and schedule automatically.

Channel coverage is broad: SMS, email, web chat, voice, phone and video, with structured AI video, voice and asynchronous interviews running 24/7 on the candidate's schedule, plus automated reminders and follow up. Availability is what is being sold, set against manual follow up that falls through the cracks. Events and schedules both start work here several times over, across six channels.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

The customer has no choice over the models, and is not told what they are. No model provider, model picker, admin setting, routing description or bring your own model option is documented.

Humanly is specific about what its AI does and silent about what runs it: conversational engagement across six channels, structured screening against rubrics and knockout criteria, ranked shortlists with transcripts and reasoning, and a stated training base of millions of candidate interactions, all described as Humanly's own capability, with conversational quality presented as the difference from keyword scanning. That is machinery the vendor operates.

Humanly's fairness and bias audit positioning addresses how the AI is governed, not which model answers. Naming a single provider would at least make the absence of choice visible; Humanly names none.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08

No interface an outside system can call is published. The site carries no developer entry, API reference, endpoint list, authentication guide, SDK or documentation site.

Two way integrations with nine named applicant tracking and HRIS platforms are prebuilt connectors Humanly maintains, not a surface a customer builds on, and the Partners page is a go to market program, not a developer one. Nothing documents an outside system calling Humanly, and directory descriptions that imply an API under the two way ATS integration are not backed by a published reference.

Sourcehumanly.io homepage, navigation and footerread 2026-09-08

Nothing lets a customer test the agent. Humanly scores every candidate consistently against rubrics and knockout criteria, but that evaluates candidates, not the agent. No sandbox, test environment, simulation, versioning, holdout or scored case before deployment is published, and nothing lets a customer try a rubric or knockout change before it runs on live applicants. That matters on a platform whose agent screens and rejects at volume without a recruiter in the loop.

Humanly positions itself on continuous third party bias auditing of its AI, and the homepage FAQ asks "How does Humanly help reduce bias in hiring". That audit covers Humanly's own models, commissioned by Humanly and published as assurance, and no route to an audit of a customer's deployed instance is published.

Sourcehumanly.io/products/ai-interviews and homepageread 2026-09-08

The agent reaches the world by talking, not by operating screens. Nothing on the homepage, the AI Recruiter page, the product navigation or the footer documents a hosted or local browser, a desktop session, or remote or local computer control. Engagement runs over SMS, email, web chat, voice, phone and video, and everything else moves through prebuilt two way connectors into nine named applicant tracking and HRIS platforms plus recruiter and hiring manager calendars.

Nothing here navigates, clicks, fills or advances anything through an interface built for a person. The candidate uses their own browser to take a video interview, which is the product delivered to a person, not the agent operating a surface, and no sourcing or capture browser extension appears anywhere in Humanly's products. Breakage when UI elements change does not apply to an SMS conversation and an ATS connector.

Sourcehumanly.io homepage and /products/ai-recruiterread 2026-09-08