ATS
Recruitee, Greenhouse, Workday
Screening evaluations mirrored back into the candidate pipeline. The recruiter reads the evidence and decides, without leaving the tool they already use.
The platform
You make the call. The engine gives you the evidence.
Who to hire, who to promote, who to move. You keep the decision. Underneath sit two intelligence engines, agents grounded in real data, and a knowledge graph that sharpens with every project. So the evidence behind your call is deterministic, traceable, and explainable.
What the core is
One engine under every people decision.
The hard part of a people decision is the evidence: do you really know what this person can do, and can you defend the call later? Two intelligence engines and a knowledge graph answer that. The same core powers screening, mobility, and skills work, so a pilot ships in weeks, not quarters, and every evaluation arrives with the evidence behind it.
Programs & agents
The Canvas, the engine room behind the work.
The Canvas is where we compose Programs and run agents against the intelligence engines. It is the engine room, not a dashboard your team logs into. Your people stay in the tool they already use, and the agents mirror their evaluations back into your ATS, your HRIS, your pipeline, where the human reads the evidence and makes the call.
The Canvas
A Skill-based hiring Program on the Canvas. A Screening Agent scores every applicant against the vacancy as applications arrive. A Culture Scorer evaluates values and working-style fit, kept distinct from technical fit.
Agent
Screening Agent
Scores every incoming candidate against the open vacancy, technical skills, experience match, qualification fit, and ranks them with the evidence as applications arrive. The recruiter still picks who moves forward.
Agent
Culture Scorer
Evaluates values and working-style fit against the team profile, kept distinct from the technical evaluation so the two signals stay legible to the person reading them.
Agent
Mobility Matcher
Reads the bench in real time and surfaces the people who fit an open project before it goes to market, so the manager can choose from inside first.
Why you can trust it
Evidence you can defend, deterministic, traceable, explainable. Not a black box.
When you sit across from a candidate, a manager, or your board, you have to stand behind the call. So the evidence under it has to hold up. Every evaluation is grounded in the intelligence engines, with the full reasoning attached, the way a sharp recruiter defends a shortlist, only at the speed of the inbox. The engine does the legwork. You make the decision.
01
Deterministic
The same candidate against the same vacancy returns the same evaluation. Reasoning is structured, not improvised, so you compare like with like.
02
Traceable
Every score links back to the evidence, the skills cited, the role data, the conversation excerpts. You can audit any line.
03
Explainable
Each evaluation comes with the reasoning in plain language, so you can defend your call to a hiring manager or a candidate.
This is the reason you can put an Everday evaluation in front of anyone and stand behind your decision. The engine is built for it from the ground up.
- 0
- intelligence engines behind your call
- 4–0 weeks
- pilot to a working engine
- 0%
- evaluations you can trace
- Deterministic
- same input, same evidence
The engines · Engine 01 / 02
Know what a person can actually do, not what a CV claims.
Skills Intelligence engine
Most people decisions start from a thin, self-reported picture of skills. Skills Intelligence is the heart of the engine, and it fixes that. It maps skills from real work data, matches them against open roles and projects, and tracks how they develop over time. The same engine powers screening, mobility, and the skill taxonomy a workforce is built on, so every call starts from the truth.
Skill extraction from work data
Reads verified skills from real artifacts, CVs, project history, performance signals, not from a self-rated checklist nobody trusts.
Role and project matching
Scores a person against an open role or project on the skills it actually needs, with the evidence behind each match so the manager can weigh it.
Skill taxonomy, automatic
Builds a current, evidence-based taxonomy from the work people do, so the map matches the territory.
Development paths
Tracks how skills evolve and surfaces the next viable move for a person, internal mobility, reskill, or stretch project.
everday · skills intelligence
Ranked shortlist
Senior Backend Engineer
A. de Vries
GoPostgresDistributed systems94%S. Bakker
GoKubernetes89%M. Janssen
Java → GoAPI design82%L. Visser
PythonData pipelines76%
Every score traces to evidence, deterministic, not a guess.
The engines · Engine 02 / 02
See your decision in the context of the market around you.
Market Intelligence engine
A good call inside the company can still miss what is happening outside it. Market Intelligence grounds your decision in the world beyond your walls, what roles look like across the sector, which skills are scarce, where the supply sits, what compensation clears. Internal skill signal meets external market signal in the same evaluation, so you decide with the full picture.
Role and sector definitions
A live read on what a role actually requires across the sector, not a stale job description from 2019.
Skill supply and scarcity
Which skills are abundant, which are scarce, and where the available people sit, by region, by industry, by seniority.
Compensation signal
Live reference points so a screening evaluation reflects what the market will actually clear at.
everday · market intelligence
Market read
Senior Backend Engineer · NL
Skill supply
Go
ScarceKubernetes
TightPython
AbundantDistributed systems
Scarce
Comp signal
€72k–€88k
Clears at
€80k
External signal, grounded in the same evaluation, not a static job description.
Agents
Agents prepare the evidence. You make the decision. The outcome teaches the engine.
Each agent does one job well, screen, match, score, summarize, and reasons over the same intelligence engines. It hands the evidence to the person making the call, never the call itself. What follows feeds back into the graph: who got hired, who moved, what landed, what did not. The engine sharpens with every project.
- 01
Run
The agent runs inside a real decision, in the tool your team already uses.
- 02
Ground
It reasons over Skills and Market Intelligence with the full evidence attached.
- 03
Hand off
A traceable evaluation lands in your existing pipeline, where the human reads it and decides.
- 04
Learn
The outcome flows back into the graph, so the engine gets sharper for the next decision.
The knowledge graph
Messy real-world data, connected understanding.
People, skills, roles, projects, decisions, and outcomes, held as one graph, cleaned and linked. We carry that structure so it does not have to be reassembled for every question. That is why your evaluations stay consistent, the evidence under each one holds together, and the engine gets sharper with every project you run.
Open by design
The evidence shows up where your team already works.
No one wants another tool to log into. The Canvas is the engine room, and the work lands where your team already works, ATS, HRIS, ticketing, chat. Connectors are open by design, so your engineering team can extend, audit, or self-host the integration surface.
HRIS
People data sync
Two-way sync against the system of record so skill, role, and mobility signal stay live.
MCP
Model Context Protocol
Expose Programs and engines to your own AI stack through MCP. Audited, scoped, revocable.
API
REST + webhooks
Run a Program from your own system. Subscribe to outcomes. Pull evidence for an audit trail.
Sharper every project
Every project makes your next decision sharper.
Every project you run adds evidence to the graph. More evidence sharpens the engines. Sharper engines ground better agents, and a clearer picture under your next decision. The engine you pilot on is the weakest it will ever be.
- 01Run a project
- 02Evidence into the graph
- 03Engine sharpens
- 04Better-grounded agents
- 05A clearer call for you
Start now, and every future call you make rides on a sharper engine.
Questions
What engineering asks first.
- No. The agent prepares the evidence, scores, matches, the reasoning behind each one, and hands it to the person making the call. You keep the decision. We remove the guesswork, not the judgment.
- No. The Canvas is the engine room, and most teams never open it. Agents mirror their evaluations back into the ATS, HRIS, or pipeline your team already works in.
- We hold your skills, roles, and history in the knowledge graph and ground every agent in two intelligence engines. So the same input returns the same evaluation, with the evidence attached, rather than a fresh answer each time. That consistency is what lets you defend the call.
- Yes. Every score links back to the evidence behind it, the skills cited, the role data, the conversation excerpts. You can pull the full trail through the API for any evaluation.
- A pilot runs in 4–8 weeks. We go into your organization, map your stack and your people process, build the agents, and leave them running on our core inside the tools you already use. You make a real call on real work at the end of it, with the evidence in front of you.