Find the Perfect Candidate by Filtering and Processing Resumes
Kognitos turns inbound resume emails into a 3-sentence summary CSV — pulling OCR and GPT-3 from a single English process.
What's in this video
A 4-minute HR-focused demo of how Kognitos automates the resume triage step that overstretched HR teams typically do by hand. The trigger is an inbound email with multiple resumes attached; the output is a CSV of 3-sentence summaries delivered straight to a recruiter.
What Kognitos does
- Email trigger: the recruiter forwards or receives resumes (or pulls from a source like Indeed) and Kognitos picks them up.
- Mixed format handling: the demo deliberately uses resumes with different layouts, colors, placements, and even typos in titles — cases that traditionally throw exceptions in legacy RPA.
- OCR + LLM under the hood: the process is written in English; Kognitos chooses OCR or NLP automatically as needed, so no developer has to pick technologies upfront.
- 3-sentence summaries: Kognitos asks the LLM to produce a concise summary for each resume. The summaries aren't templated — each one surfaces the candidate's most relevant experience, education, and stats.
- CSV delivery: the run packages summaries into a CSV emailed to the recruiter. The output can also be the email body, a different file format, or pushed directly into an HRIS — recruiter's choice.
Why HR teams care
HR is often overloaded with high turnover, heavy resume volume, and onboarding/offboarding work. Kognitos lets a recruiter define the summarisation brief in English (and tighten it later if they want specific fields) without waiting for an RPA developer or NLP specialist. The same English process can handle resumes from Indeed, from email, from a referral inbox — anywhere they arrive.