Talyntry
A full AI recruiting platform, built from scratch
An applicant tracking system for tech staffing — AI resume parsing, a ten-stage pipeline, candidate matching and team collaboration. Product, brand and landing page, all of it.
Maria Kowalski
Sr. Cloud Architect · Senior IC
Vertical
Cloud / DevOps
Work auth
US Citizen
Rate
$92/hr
Owner
Andrew Silva
Pipeline progress
- 194%
Maria Kowalski
- 291%
Priya Patel
- 390%
Lisa Torres
- 489%
Sofia Marchetti
- 588%
James Sullivan
3s
avg. resume parse
10
pipeline stages
Top 5
AI matches per role
Real-time
team sync
A working miniature of the real pipeline — click a stage to advance a candidate, or run the matcher.
Published product figures
68%
Faster time to fill
94%
Placement retention
3s
Avg. resume parse
10
Pipeline stages tracked
The situation
Recruiting runs on tooling that is either enterprise-priced and bloated, or a spreadsheet with ambitions. Small and mid-size staffing firms end up losing candidates in email threads and rebuilding the same pipeline in a different tool every eighteen months.
Talyntry was built to be the thing in between: an applicant tracking system that a three-person desk can run on day one and a growing firm does not outgrow. It targets tech staffing specifically — telecom, cybersecurity, AI and ML, cloud and DevOps, and executive placement.
The differentiator is where the AI sits. Not a chatbot bolted onto a CRM, but AI doing the two jobs recruiters actually lose hours to: reading resumes and working out who in the bench fits the role that just came in.
What was built
- 01
AI resume parsing
PDF, DOCX and image resumes parsed into structured candidate records — name, title, skills, work authorisation, seniority — in about three seconds. DOCX is converted to text before it reaches the model, since the API only accepts PDF as a document type.
- 02
Ten-stage pipeline with drag and drop
Sourced through Placed, plus On Hold and Rejected, as a Kanban board with full drag-and-drop between columns and a colour map that holds across every view in the app.
- 03
AI candidate matching
Scans the entire pipeline against a job order and surfaces the top five with confidence scores, so a recruiter opens a new req against a ranked bench rather than a blank search.
- 04
Team collaboration and audit trail
Shared workspace with notes, interview scorecards and a full activity timeline — stage changes, resume uploads, submissions, edits — with real-time sync across the team.
- 05
Job orders and AI job descriptions
Create and assign roles, track submissions per req, and generate job descriptions from minimal input.
- 06
Brand and landing page
Identity, logo system and a one-page marketing site with a cursor-reactive dashboard mockup, pricing tiers and a security section covering encryption, role-based access and GDPR.
Where it landed
Talyntry runs in production on Vercel with Supabase behind it, serving live recruiting desks. Resume parsing, candidate matching and job description generation all run through the Anthropic API.
The interactive panel above is a working miniature of the real pipeline — the same ten stages, the same colour map, the same matching behaviour. Click a stage to advance a candidate, or run the matcher.
More work
Find Out What You’re Missing
Thirty minutes. We’ll find where the work is leaking — unsold estimates, response time, or a database nobody has touched in three years — and what it would cost to fix. If it isn’t worth it yet, we’ll tell you that instead.
No sequence, no newsletter. One reply from the person who’d build it.