
Case study
OS for Recruiters
A flexible recruiting workspace with AI inside the work
at SeekOut ↗Choose your depth
The problem, my contribution, outcomes, and strongest screens.
My role
Product design · Product strategy · AI system design
Product focus
Reframing a linear recruiting product as a flexible, chat-first workspace built around requisitions and structured artifacts.
The Short Version
SPOT v1 proved that recruiters would use AI in a live search. It also exposed the limits of a product built as a sequence of stages. Requirements changed after intake, calibration happened once real profiles appeared, and hiring-manager feedback arrived while candidates were already moving. For v2, I helped rethink SPOT as a workspace. Each requisition has its own chat, context, and structured artifacts for intake, market analysis, sourcing, engagement, reports, and settings. Recruiters choose where to work. The AI can research, draft, score, and propose changes without turning its preferred sequence into the process.
The Mess
The hardest problem was trust. Recruiters reported false positives in scoring and rewrote outreach that missed their bar. Their approvals, rejections, and edits never reached the model, so the tenth search started with the same blind spots as the first. The interface added another cost: running an evaluation, changing a status, or finding one candidate often meant crossing several screens. People moved work into Claude, personal notes, and Excel because those tools were faster. That left SPOT with an incomplete record and gave its AI less context for the next decision.
My Part in This
- Helped move SPOT from a prescribed stage flow to a chat-first workspace that can absorb changes during a search.
- Defined the requisition as an isolated unit of work, with its own artifacts, conversation history, edits, and AI context.
- Designed market research, talent-pool analysis, constraint testing, and calibration as one place to challenge a brief before sourcing at scale.
What Got Better
- Defined a workspace model that lets recruiters change direction without losing the history or criteria behind a search.
- Made artifacts the source of truth, giving chat a stable set of job details, strategies, candidates, messages, and reports to work from.
- Set clear approval boundaries for AI research, drafting, scoring, outreach, and hiring-manager sharing.
- Connected day-to-day candidate work with a cross-requisition task queue, recovery states, and an audit trail.
Impact
What Actually Moved
For the Business
Defined one requisition model for chat, artifacts, candidate work, reporting, and audit history, giving product teams a shared structure for SPOT v2.
For the Humans
Recruiters can change direction, inspect AI proposals, and return to the current state of a search without reconstructing it from chat or personal notes.
For the Spreadsheet
The product brief sets measurable targets, including outreach within two hours of intake, a 20% reply rate, and 80% recruiter approval of outreach-ready candidates.
Continue reading