Executive Recruitment & Talent Intelligence Consultant

About PeerSearch.ai

PeerSearch.ai is an executive talent intelligence platform built around a single idea: the fastest way to map a market is to start from a leader you already trust. Paste one executive's profile link and the platform returns that person's employer, title and photo first, then streams up to 200 comparable profiles in real time, grouped by function and ordered by seniority.

Recruiters and talent acquisition teams then narrow that map in plain language — "VPs of finance with P&L ownership", "operators with international experience" — save the people who matter into projects that persist across searches, and export a board-ready PDF or an Excel file for their own tracking.

We are a small team. That means the person in this role has direct influence over how the product works, not just how it is used.

The role

We are looking for a senior practitioner. You have spent a decade or more inside executive search or talent acquisition — building talent maps, benchmarking leadership teams, defending a shortlist in front of a client, and owning the research function behind it. You know what makes a market map credible and, just as importantly, what makes one useless.

You have also moved past the point where tools are a black box. You understand where profile data comes from, how it degrades, and how an AI system decides what to return — and you are not willing to put your name to output you cannot defend.

In this role you bring that judgement into the product. You run real searches for real briefs, you set the bar for data quality and for what the AI is allowed to say, and you tell us precisely where the platform falls short of what a professional search would stand behind. You will be the strongest critic the platform has.

What you'll do

Run and shape real searches

Take live executive search and talent acquisition briefs and work them end to end: anchor on a benchmark leader, map comparable talent, narrow with prompts, and deliver a shortlist you would sign your name to.

Set the standard for what counts as a credible market map here, and hold every deliverable to it — including your own.

Produce market maps and competitor leadership views for clients and prospects, and present them in a way that holds up in a partner or hiring-manager meeting.

Judge whether the profiles returned are genuinely comparable — function, seniority, scope, tenure — and flag every case where they are not.

Define the talent intelligence process

Document the process the platform is meant to encode: how a search should start, what a complete talent profile contains, how results should be grouped and ranked, and what a good executive summary of a mapped market actually says.

Turn client feedback and your own experience into clear, specific requests the team can build against — not vague wishes.

Test each release against real briefs and report what improved, what regressed and what is still missing.

Own data quality, coverage and structure

Specify the profile schema: every field, its definition, its source, its refresh interval, its validation rule, and what happens when the field is missing or cannot be verified.

Design and run the quality audit. Sample a market, score accuracy, duplication, staleness and coverage, quantify the error rate, and separate systematic failure from noise.

Own entity resolution: company-name variants, subsidiaries and rebrands, duplicate people, title inflation. Define how a merge is verified rather than assumed, because silent merges are where trust dies.

Build the measurement — a repeatable scorecard the team can watch — so a release is judged on evidence rather than impression.

Say plainly when a market is too thin or too stale to map reliably, and what would have to change for it to be mappable.

Evaluate and push on the AI

Build and maintain the evaluation set: a fixed library of briefs and their known-good answers, against which every prompt change, ranking change and model change is scored.

Score shortlist quality the way a search professional would — precision, recall, and whether the top of the list is the list you would actually send.

Debug failure modes precisely: a fabricated profile, a plausible-but-wrong summary, a prompt that quietly drifts, a ranking that favours the wrong signal. Name the field or the check that would have caught it.

Push on retrieval and matching. Tell us when semantic similarity is picking up the wrong thing, what should replace it, and how we would know the change worked.

Decide what belongs in the product and what does not — where a model genuinely saves a researcher time, and where it just produces confident nonsense.

Visualization and presentation

Decide what a reader needs to see first. You can look at 200 profiles and know which grouping, ordering and summary language makes the market legible in thirty seconds.

Shape the on-screen map — function grouping, seniority ordering, the executive summary and its clickable narrowing phrases — so each click produces a genuinely tight shortlist.

Design the deliverable: what belongs on the one-page summary grid (name, firm, current title, one-line summary) versus a full bio page per person, and how the Excel export should be structured for someone who will sort and filter it.

Critique density and clutter. Say what to remove, and be right.

Enablement and content

Build the reference material: example searches, prompts that consistently produce sharp shortlists, and walk-throughs that show a new user the value in their first session.

Support prospective and current users with product guidance and honest answers about what the platform does and does not do.

What you bring

Search judgement

Practitioner depth — 10+ years in executive search, talent acquisition, talent intelligence or market research — agency, in-house or a mix — including time leading a research or analytics function.

Hand-built maps — You have personally built talent maps, succession benches and competitor leadership views, and you have been accountable for one in front of a client.

Tool fluency — Deep fluency with sourcing tools and talent intelligence or market data platforms, and a precise view of where each one falls short.

Clear writing — You write tight, specific, client-ready prose. No filler, no hedging.

Data

Query fluency — SQL is a working tool for you, not a phrase on a CV — joins, aggregations and window functions across hundreds of thousands of profile rows.

Schema design — You have specified a data schema or field dictionary: definitions, sources, allowed values, validation rules, and how missing or unverified data is handled.

Measured QA — You can design a quality audit and quantify the result — sample size, error rate, coverage by industry, firm size and geography — and say what the number means.

Dedup — Entity resolution and deduplication do not surprise you. You know merges fail silently and you know how to catch it.

Data instincts — You are comfortable with large structured datasets and can tell when a distribution looks wrong before anyone checks it.

Modelling — You can move from a question to a spreadsheet or notebook model that answers it the same afternoon.

AI

Daily practitioner — You use AI assistants every day for real research work, and you know exactly where they fail quietly.

Prompt craft — Prompting is a craft for you: structure, constraints, worked examples, and making an output repeatable rather than lucky.

Evaluation — You have evaluated model output against ground truth — a test set, precision and recall on a shortlist, or two prompt versions scored on the same fifty profiles.

Retrieval literacy — You understand retrieval well enough to explain embeddings, semantic matching and why a particular query returned what it did.

Workflow building — You have built agent or automation workflows around model output rather than one-off prompts, and you have automated part of your own research process.

Scepticism — You can tell the difference between a genuine AI capability and a marketing claim, and you say so out loud.

Working style

You own an ambiguous problem end to end in a small company without waiting for a brief.

You disagree with evidence, and you are comfortable being overruled by a number.

Even better if you have

Workforce planning, org design or succession experience at enterprise scale.

You have led or mentored a research team, or run a research quality function.

HR technology implementation or recruiting-stack administration experience.

You have published evaluation methodology, or written publicly about how you test AI output.

Python or R for analysis beyond SQL.

A network in executive search or talent acquisition willing to tell you what they actually need.

How we work

Remote-first and asynchronous. We care about the quality of your output, not your hours online.

Small team, short feedback loop: what you flag today is typically triaged this week.

Discretion matters. You will see client briefs and confidential searches, and you will keep them that way.

We build in the open internally and are careful about what we say externally. The way profiles are sourced is not discussed outside the team.

Compensation and engagement

This can be structured as a full-time role or a part-time contract with a defined monthly scope. Rate or salary is set from your experience and the scope we agree — tell us what you are looking for and we will be direct about what we can offer.