Dweet Lever Integration - What It Does and When to Use It

The Dweet Lever integration connects Nova's automated candidate reviews to your Lever ATS so hiring teams can prioritize applicants without leaving their hiring workflow. It is implemented as a one-way feed: Nova reads applications, produces standardized feedback scores, and inserts those scores into Lever candidate opportunities. This article explains exactly what the integration does, how it works in practice, when it produces the most value, and what operational checks you should run before and after enabling it.

At a high level, the integration automates the initial screening step: Nova evaluates resumes and application data against your configured criteria, then Dweet writes the resulting review metadata (scores, tags, brief notes) into Lever. The result is consistent, auditable screening data attached to each candidate record that recruiters and hiring managers can filter and sort on directly in the ATS.

Key features of the Dweet — Nova — Lever flow

• One-way review feed Nova pushes evaluation results into Lever; Lever records are updated with scores and short comments so the ATS remains the single source of truth.
• Standardized scoring Candidates receive consistent numeric or categorical scores based on the same rubric, reducing subjectivity for the first-pass screen.
• No context switching Recruiters continue working in Lever while seeing Nova insights inline, removing the need to switch tabs or tools during triage.
• Priority-driven workflow Score-based sorting lets teams focus on high-fit applications first, improving interview-to-offer yield when used correctly.
• Audit trail Each automated review is logged so teams can trace why a candidate was deprioritized or advanced.

How it works technically: Nova uses your configured evaluation models to inspect application fields and uploaded resumes; Dweet acts as a lightweight integration layer that maps Nova output to Lever opportunity fields using Lever's API. When a new application arrives, Nova produces a review object; Dweet authenticates to Lever and writes the score, tag, and a short comment to the candidate opportunity as a custom field or note.

Before vs After: Screening with and without the integration

When to use the Dweet Lever Integration

• High-volume requisitions Use it when roles receive hundreds of applicants — the integration filters noise and surfaces the top percentile for human review.
• Clear, repeatable screen criteria Best when you can define objective signals (skills, years of experience, certifications) that automation can score reliably.
• Short time-to-hire goals When speed matters, automatic prioritization reduces time spent on low-fit candidates and accelerates pipeline conversion.
• Need for consistent first-pass screening If manager expectations demand consistent, reproducible screening across recruiters, the integration enforces a single baseline.
• Desire to minimize tool sprawl If recruiters prefer staying in Lever, this integration centralizes screening insights within the ATS rather than requiring separate dashboards.

Limitations and cases to avoid: the integration is a one-way feed that writes Nova's reviews into Lever; it does not replace human judgment and should not be used as the sole gatekeeper for high-stakes hires. The model's performance depends on clean job descriptions and consistent input data — poorly structured application forms or empty resumes reduce accuracy.

Common questions and answers

Best practices for setup and adoption

• Map fields intentionally Decide which Nova outputs should be visible in Lever (numeric score, tags, short note) and map them to clear, documented custom fields.
• Pilot with a single team Start with one hiring team or requisition type to validate thresholds and calibrate the model before full rollout.
• Set clear triage rules Establish what score thresholds mean operationally (e.g., score > 80 = auto-priority, 50–79 = recruiter review).
• Log decisions Ask recruiters to add a brief reason if they override automation to build a feedback set for future tuning.
• Monitor model drift Review precision/recall and false positive rates periodically; update criteria when role requirements change.

Workflow examples: For a high-volume customer support role, Nova can screen for specific keywords, experience years, and certifications then push a pass/fail or score to Lever; recruiters focus only on candidates above the pass threshold. For niche technical roles, use Nova to surface candidates who match core technical skills and relevant project descriptions, while keeping final technical validation with engineers.

Operational metrics to track after enabling integration

Implementation checklist: create a scoped API user in Lever, identify which Nova fields to push, configure Dweet mapping, run a dry-run on historical applications for calibration, pilot live traffic with a single team, collect override feedback, and then roll out incrementally with monitoring dashboards in place.

Troubleshooting tips

Estimating ROI — a quick example: if one recruiter spends 4 hours per day on initial screens and automation reduces that by 75%, the team saves 3 hours per recruiter per day. Multiply by recruiter hourly cost and number of recruiters to produce a monthly time-savings figure, then compare to integration and subscription costs to determine payback period.