Talent Pool: Conversational search for your talent pool
You can now search your talent pool through a conversation. Tell Nova what you’re looking for, answer a few calibration questions, and it finds past applicants who match, going beyond simple keyword matching to surface candidates based on real experience and fit. Refine results with follow-up messages and build a shortlist as you go.AI Outbound Talent Search: Email notification when search completes
You now receive an email when your talent search finishes its first round of results. The email includes a summary of how many candidates were shortlisted by fit level and a direct link to view results. No need to keep the page open during long-running searches.Criteria Builder: Suggested actions after each response
The criteria builder agent now shows clickable action buttons after each response. Instead of typing your choice, click a suggested action to continue the conversation. The buttons match the options the agent offers in its message, so you can move through the setup flow with a single click. Your selected action appears as a pill in the chat for context.Fraud Detection: Fewer false positives
Fraud detection now produces significantly fewer false alarms on legitimate candidates. Improvements to how the system interprets common consumer email addresses, career transitions, and employer claims reduce false positives by roughly 40% at the recommended operating threshold, while maintaining the same overall detection accuracy.AI Outbound Talent Search: Search history on the setup page
Recent searches now appear below the setup form so you can jump back into any previous session with one click. Each row shows the search title, how long ago it ran, and whether it completed.AI Outbound Talent Search: Cleaner setup page for ATS-free accounts
If your account is not connected to an ATS, the setup page now shows a single “Paste job description” form with no extra tabs. The page heading and layout are also refreshed to feel less like a prototype and more like a finished product.Criteria Builder: Smoother criteria generation
Criteria generation now happens inside the conversation. Nova asks targeted questions first, then generates criteria based on your answers, so the whole flow feels like one continuous chat. If you mention specific brands or companies, they are preserved in the generated criteria.Criteria Builder: Verifiability warnings
Nova now flags criteria that are difficult to verify from resumes alone, like budget ownership or process governance. You can adjust these before scoring starts, or move them to interview checks.Criteria Builder: More role-specific criteria titles
Criteria titles now pick up more of the specific language from your job description, so they read naturally alongside the role you are hiring for.Criteria Builder: Live progress while thinking
You now see a live timer and word count while Nova is working, so you always know where it is in the process.Criteria Builder: Easier-to-read calibration results
Calibration summaries now read as plain-language analysis, making them quicker to scan and act on.Scoring: Stronger candidates score higher
Scoring now better recognizes candidates who clearly match your criteria, so strong fits receive scores that reflect their strength.AI Outbound Talent Search: Watch it work end to end
See the full outbound sourcing flow in action: intake calibration, multi-angle talent search, criteria-aligned scoring, and a ranked shortlist with receptivity signals.
AI Outbound Talent Search: Live progress during candidate evaluation
You now see real-time progress while the agent evaluates candidates and expands searches. Instead of a static spinner, you see live counters like “Scored 5/23” during candidate evaluation and “Processed 3/5 seeds” during search expansion. Progress updates in place and disappears when the step completes, leaving a clean summary.Onboarding: Try Nova without connecting an ATS
You can now skip the ATS connection step during onboarding and go straight to AI Outbound Sourcing. Click “Continue without an ATS” on the ATS selection page to start sourcing candidates by pasting a job description. You get 2 trial searches per month. Pages that require ATS data (Jobs, Candidates, Auto-Tagging) show a clear prompt to connect your ATS when you are ready.AI Outbound Talent Search: Give your search a head start with intake context
When you start a new search, you now see a prompt to paste hiring manager notes, intake call transcripts, or other context that goes beyond the job description. Click a suggestion to get a template, or paste your own notes directly. The agent uses this context to ask sharper calibration questions and build a more targeted search from the start. You can skip it if you prefer to jump straight in.AI Outbound Talent Search: Rate candidate assessments
You can now give quick feedback on each candidate’s fit rating directly from the shortlist. Thumbs up confirms the assessment is accurate. Thumbs down opens a short form where you pick your own fit level and explain why you disagree. Your feedback is saved per candidate and included in CSV exports, helping us measure and improve search quality over time.AI Outbound Talent Search: Collaborative intake before searching
The agent now asks targeted clarifying questions before starting a search when the job description is ambiguous or missing key details. Instead of guessing, it checks priorities like must-have experience, flexibility on location, or deal-breakers, then uses your answers to build a sharper search. You can also paste a job description without a linked job posting and specify the company name manually.AI Outbound Talent Search: Receptivity scoring
Candidate assessments now factor in how likely someone is to be receptive to outreach. The agent looks at signals like recent role changes, short tenure, lateral moves, and career trajectory to flag candidates who may be more or less open to a conversation. This surfaces as part of the scoring rationale, not as a separate score.AI Outbound Talent Search: Faster shortlist filtering
The shortlist sidebar now includes one-click fit chips (Best Fit, Strong Fit, Consider) and a minimum score filter. You can quickly narrow large shortlists to the candidate tier you want without scrolling through every profile.AI Outbound Talent Search: Clearer candidate cards and richer CSV export
Candidate cards now prioritize a concise summary and structured work history (title, company, and date range), with expandable experience for deeper review. CSV exports now include current and previous role/company columns so handoff and offline review are faster.AI Outbound Talent Search: Conversational search with follow-up messages
Talent search now works like a conversation. After the agent delivers its first batch of candidates, you can send follow-up messages to refine results: ask for more candidates from a specific industry, adjust your requirements, or explore a different direction. The agent picks up where it left off with full context from prior rounds.Each checkpoint now ends with clear, actionable questions. The agent recommends an answer and explains what changes in the next round, so you can confirm with a single click instead of thinking through trade-offs from scratch.AI Outbound Talent Search: Smarter scoring for hard-to-evidence criteria
The agent now recognizes when a requirement (like budget ownership or process governance) is unlikely to appear on LinkedIn profiles. Instead of penalizing candidates for missing information, it flags these as phone-screen verification items and adjusts scoring accordingly, so strong candidates are not filtered out before you can talk to them.Talent Pool: ATS context in candidate search explanations
Talent Pool search now uses ATS context in candidate explanations, including application question answers, recruiter notes, activity history, and rejection details when available. This helps you spot outreach risks earlier, such as salary expectations, notice period, sponsorship needs, or prior interview feedback.Available for Teamtailor, Greenhouse, Ashby, Workable, and Pinpoint.Fraud Detection: Stronger LinkedIn reasoning and safer fallback behavior
Fraud Detection now uses richer LinkedIn profile context when available, including structured profile sections and timeline details, to improve reasoning quality on identity coherence, claim consistency, and credential traceability.Fraud Detection: Filter by job posting
You can now filter the Fraud Detection dashboard by specific job postings. Use the new “Jobs” dropdown to select one or more roles and see summary cards, trend charts, and the assessment table scoped to those jobs. The dropdown is searchable by job title and ID, and shows the number of assessments per role.Fraud Detection: Customer-facing dashboard
You can now review fraud risk assessments directly in Nova. The new Fraud Detection page shows a summary of all screened applications, broken down by risk level (low, medium, high) with visual trends over time.Key features: date range filtering, clickable risk cards to filter the table, a detail dialog for each assessment with the full analysis, and charts showing how risk rates and assessment volumes are changing. All metrics include plain-language explanations so you can interpret results without technical context.The page is only visible when Fraud Detection is enabled for your company.
Criteria Builder: Understand why candidates received their score
You can now paste a candidate’s ATS URL into the Criteria Builder chat to see exactly why they received their Nova score. The agent will show you the score breakdown, highlight which criteria influenced the result, and help you refine your criteria if something looks off.Fraud Detection: Cross-candidate matching on phone and LinkedIn
Cross-candidate duplicate detection now checks phone numbers and LinkedIn profiles in addition to email addresses. This catches a specific fraud pattern: the same person submitting multiple applications under different names. When the same phone number or LinkedIn profile appears on applications from “different” candidates, that is a strong signal of coordinated fraud. Legitimate candidates reapplying under their own name are not affected.In our benchmark, shared contact details appear almost exclusively on fraudulent applications, making this one of the highest-precision signals in the system. Nova uses structured contact data from your ATS as the primary source, with resume text extraction as a fallback for email and LinkedIn.Fraud Detection: Recruiter-friendly risk notes
Fraud risk notes now avoid internal terminology and use clearer language that fits ATS workflows. This makes the analysis easier to review quickly.Notes are now easier to skim with a consistent structure: risk, takeaway, top reasons, and optional reassuring signs. Top reasons now use direct factual wording instead of internal process phrasing.
Fraud Detection: Hidden content and keyword stuffing detection
Nova now detects hidden text embedded in resume PDFs. Some fraudulent resumes hide invisible keyword lists using tiny text or overlapping duplicate text layers to game ATS keyword matching systems. Nova’s PDF forensic analysis identifies these manipulation techniques and factors them into the fraud risk assessment.AI Talent Search: Usage limits
AI talent searches are now limited to 5 searches per 30-day rolling window per company (default). The search button shows your current usage (e.g. 3/5), and you will see a clear message when the limit is reached. Failed searches do not count against your limit. Contact support to adjust your limit.Jobs Dashboard: Cleaner, denser job list
The jobs table now uses a two-line row layout inspired by deployment-list patterns. Each row shows the job title, Nova pipeline status, job status, and relative timestamp on the first line, with the job ID, location, criteria status, and scoring progress on the second line. All row actions are consolidated into a single overflow menu, and sorting is controlled from a toolbar dropdown instead of column headers.The result is a significantly narrower table that fits comfortably without horizontal overflow, while keeping the most important information scannable at a glance.Fraud Detection: Multi-layered resume fraud analysis
Nova’s fraud detection system is now production-ready with a major upgrade to its analysis capabilities. Instead of relying on resume text alone, Nova now combines text analysis with multiple independent external validation signals to produce significantly more accurate fraud assessments.What it catches: Nova analyzes resumes across multiple dimensions: timeline integrity, career progression plausibility, skill claim credibility, document metadata anomalies, email and identity verification, LinkedIn profile validation, and cross-candidate duplicate detection. When multiple independent signals converge, fraud probability increases accordingly. When only surface-level style concerns are present, Nova keeps risk scores low.Benchmark-validated accuracy: We built and maintain an internal benchmark of hundreds of real resumes (both confirmed fraud and verified legitimate applications) sourced from production hiring pipelines. At the recommended operating point:- Over 93% precision: when Nova flags a resume, it is correct over 93% of the time
- Very low false alarm rate: fewer than 1 in 25 legitimate resumes are incorrectly flagged
- High-confidence mode: at the highest confidence level, Nova has achieved zero false positives across the full benchmark
Talent Pool: Compact search results layout
The search results page now uses a denser layout so the candidate table gets significantly more vertical space. The header, filters, and feedback nudge are compressed into a single toolbar row, and table cell padding is tighter. On a typical laptop screen you should see roughly two more candidate rows without scrolling.Criteria Builder: Calibration matrix view
You can now open a table view of your calibration results directly from the criteria builder. Click “Table View” after calibration completes to see all candidates and criteria in a grid, making it easy to spot patterns across your sample at a glance. Click any candidate name to view their full assessment details.Scoring: Smarter must-have evaluation
Nova is now better at reading between the lines on must-have requirements. When the resume already contains enough information to assess a must-have, Nova uses that evidence directly instead of suggesting an interview check. For example, if a role requires 5+ years of experience and the resume shows 3 years, Nova will reflect that gap in the score rather than treating it as an open question.Interview verification is now reserved for must-haves where the resume is genuinely silent, like work authorization or a specific tool that simply is not mentioned.Scoring is also more role-specific. A strong individual contributor applying to a management role is now evaluated against the management requirements, so the score reflects fit for the actual role rather than general career strength.Criteria Builder: More stable calibration re-runs
When you re-run calibration with the same candidates, Nova now keeps prior evaluations for criteria you did not change. This makes score changes more predictable and focused on the criteria you edited, so loosening one requirement does not cause unrelated score swings.Scoring: Better handling of missing resume details
Scoring now treats missing resume detail as uncertainty instead of an automatic failure. When a must-have is plausible but not explicit, Nova marks it as partial and suggests targeted interview checks, which helps you keep strong candidates in play while still protecting shortlist quality.You should also see more consistent shortlist decisions at the top end of scores. Nova is now stricter about promoting candidates when critical must-have evidence is still unresolved, and it gives clearer verification-first interview guidance for those cases.Talent Pool: Simpler and more reliable indexing pipeline
Talent Pool indexing now runs on a simplified discovery and indexing flow with inline resume extraction. This removes legacy batch queue complexity, reduces background task overhead, and keeps indexing output and enrichment behavior consistent.Talent Pool analytics now label this state as Pending and define it as eligible candidates waiting to be processed, replacing queue-specific wording.Talent Pool: Filter by allowed job postings
You can now select specific job postings to include in your talent pool. When any job postings are selected, only candidates who applied to at least one of those jobs will be indexed. Leave the list empty to include all job postings (the default). This works alongside the existing blocked job postings filter to give you precise control over which candidates enter the talent pool.Calibration: Evidence-based feedback from real resumes
Integrations: High potential stage automation
You can now automatically advance high-scoring candidates to a “High Potential” stage, just like the existing low potential automation. Configure both automations together to create a complete triage system: low-scoring candidates move to one stage, high-scoring candidates to another, and middle-range candidates stay in New Applications for manual review.Each automation can be toggled independently, and a visual preview shows exactly which score ranges map to which stages. This feature is available for Lever, Greenhouse, and Nova ATS.Talent Pool: View indexing analytics and progress
You can now see detailed analytics for your talent pool indexing. Visit the new Analytics page from your Talent Pool to view total indexed candidates, success rates, queue health, and indexing progress per job. A time-series chart shows indexing activity over time, helping you understand your talent pool’s growth.Criteria Builder: Improved error handling and retry reliability
Fixed an issue where the Criteria Builder could get stuck on the loading screen if the background task failed. You’ll now see a clear error message with a “Retry Initialization” button that works reliably.Criteria Builder: Add intake context before question generation
You can now paste intake call notes, transcripts, or hiring manager requirements before we generate clarifying questions. This helps us ask better questions and prioritize criteria based on your specific context.
Calibration: Test criteria with your own resumes
When calibrating your criteria in the chat, you can now upload your own resume files to include in the test alongside ATS candidates. This helps you verify your criteria work for specific candidate profiles you have in mind.Drag and drop up to 5 resume files (PDF, DOC, DOCX), and the calibration button shows exactly what will be tested: “Start Calibration (5 ATS + 3 uploaded)”. Your uploads and results persist if you refresh the page.Criteria Builder: Clear next steps when you’re ready
When your criteria are ready, Nova now offers two clear options: start screening candidates immediately, or test your criteria with calibration first. Just say “yes” to start Nova or “let’s calibrate” to test things out. Nova handles the navigation and setup for you.
Criteria Builder: Edit criteria inline
You can now edit criteria text and importance directly without opening a separate edit mode. Click on the importance button to change it instantly, or click into the text to make changes. Everything saves automatically when you click away.Scoring: Quick feedback on Nova’s scores
You can now give quick feedback on Nova’s scores directly from your ATS. Each scored candidate note includes “Agree” and “Disagree” links that let you tell us whether the score felt right.Agree: One click confirms the score was accurate.Disagree: Opens a short form where you can optionally explain what was off.Your feedback helps Nova learn and improve scoring accuracy for your team over time. All feedback is tied to specific scoring runs, so we can track exactly which version of the criteria and scoring logic you’re responding to.Criteria Builder: Leaner, smarter criteria
Nova now generates concise criteria that trust Nova’s scoring system’s intelligence. Instead of spelling out obvious distinctions or over-explaining requirements, criteria focus on what matters.Before: Verbose criteria with redundant explanationsCollection Merchandising ExperienceAfter: Lean titles with structure only where needed
Experience working with fashion collections from concept to completion. Must have worked on seasonal RTW collections at a luxury fashion house. Should understand pre-production research, production optimization, and post-production editing processes.
Collection Merchandising, Buying, or Styling ExperienceNova’s scoring system understands that “luxury fashion house” has specific expectations, that “collection merchandising” differs from store-level visual merchandising, and that “RTW” means ready-to-wear. Structure is now reserved for non-obvious screening information. For example, the exclusion above clarifies we need HQ-level collection work, not retail floor experience.Smarter consolidation: Nova now automatically merges overlapping requirements. Instead of generating separate criteria for “fashion retail experience,” “luxury brand background,” and “sustainable fashion knowledge,” it consolidates them into a focused criterion like “Experience in Fashion, Luxury, or Department Store Retail” with a simple note: “Include: sustainable or eco-conscious fashion brands.”
Exclude: retail store management or visual merchandising-only roles
Criteria Builder: More natural setup questions
The questions you answer when setting up job criteria now feel more like a conversation with a recruiter. Instead of technical configuration options, you’ll be asked about your priorities, flexibility, and deal-breakers in plain language.Before:“What level of experience is required?”After:
- Entry level (0-2 years)
- Mid level (3-5 years)
- Senior level (5+ years)
- Principal/Staff level required
“What’s the ONE thing you can’t compromise on for this role?”You’ll also be asked about flexibility (“Would you consider candidates from adjacent industries?”) and deal-breakers (“What would make you immediately pass?”). Your answers directly shape how strict or flexible the generated criteria are.
- Strong technical skills in the core stack
- Previous experience in our industry
- Leadership or mentorship ability
- Startup or high-growth company background
Criteria Builder: Smarter conversations with personalized guidance
Nova’s criteria builder agent now gives you more helpful first interactions. When you open the criteria builder, Nova scans your criteria and may highlight a quick improvement opportunity—like a requirement that’s hard to verify from resumes. You’ll see something like:“Your criteria are set up. I noticed the ‘work authorization’ requirement might be tricky to verify from resumes alone. I can help refine that, or work on whatever else you have in mind.”When criteria look solid, Nova offers general guidance instead of forcing suggestions.
Criteria Builder: Expert recommendations when you need them
When Nova suggests multiple options for a criterion, it now proactively recommends one with clear reasoning:“Would one of these work?This saves you time while still letting you choose a different option.I’d recommend the first option because it’s specific enough to be meaningful but general enough to capture candidates using either database. Want me to add it, or would you prefer the other?”
- PostgreSQL or MySQL experience (Preferred)
- General SQL database experience (Preferred)
Criteria Builder: Recover from errors without losing progress
If something goes wrong during your conversation (network issues, service interruptions), you’ll now see a clear error message with a Retry button. Click it to resume right where you left off—no need to start over or refresh the page.Settings: Resume-only scoring mode
You can now tell Nova to score candidates based solely on their resume, ignoring application question answers entirely.Why this matters: Application answers are increasingly written by AI, which can inflate scores for candidates who haven’t actually demonstrated those skills on their resume. When you need to verify real experience—not polished claims—this setting ensures Nova evaluates what candidates have actually done, not what they say they can do.Enable “Exclude Application Questions from Scoring” in Company Settings when you want stricter, resume-first evaluations.Coming soon: Per-criterion control over evidence sources. Set individual criteria to evaluate from “resume only”, “application answers only”, or “both”—so you can be strict about verifiable experience while still assessing writing quality from application responses.Criteria Library: Save and reuse criteria across jobs
Stop rewriting the same criteria for every role. Save criteria to your company library and reuse them across any job with a click. Perfect for company-wide standards like tenure requirements, startup experience, or communication skills.Thanks to Mike Bradshaw for the feature request!Criteria Builder: Auto-sorted criteria and streamlined editing
Criteria now automatically sort by importance: Must Have criteria appear first, followed by Preferred, then Nice to Have. This applies everywhere including the criteria panel, the library dialog, and when editing. We’ve also added quick actions to add new criteria or pull from your library directly from the criteria panel header.Criteria Builder: Review multiple changes at once
When Nova suggests improvements to your criteria, you now see all proposed changes together in a single card. Each change shows exactly what will be modified with a clear before-and-after diff, making it easy to understand the impact. Approve all changes with one click, or reject and provide feedback to guide Nova’s next suggestion.
Talent Pool: Find candidates from adjacent career paths
Talent pool search now surfaces candidates from related career paths who might be great fits for your role. For example, when searching for a Marketing Coordinator, you’ll now see relevant candidates with backgrounds in Communications, PR, or Content roles. This helps you discover qualified talent you might otherwise miss.Application fraud detection (beta)
Nova now automatically analyzes highly-scored applications for signs of fraud, flagging timeline inconsistencies, impossible technology claims, and other suspicious patterns directly in the ATS note.Criteria Builder V2: Conversational criteria creation
Build screening criteria through natural conversation instead of filling out forms. Ask Nova to generate criteria for your role, answer clarifying questions to fine-tune them, and request changes by simply describing what you want. Your conversation is saved automatically, so you can pick up right where you left off.Jobs: Smoother search filter clearing
Clearing the search filter in the jobs table now works reliably without flickering. Previously, the filter value could briefly reappear when clearing it quickly.Analytics: Date range selection for monthly charts
The Applications and Candidates Created Per Month charts now include a range selector. Drag the handles to focus on specific time periods and see the total count for your selected range.Jobs: Faster loading for companies with many jobs
The Jobs page now loads much faster for companies with large job catalogs.Ashby Integration: Improved onboarding with feedback form setup
The Ashby integration now includes a guided setup flow that helps you configure how Nova records candidate assessments. During onboarding, you can choose whether to use feedback forms or notes based on your email notification preferences. This ensures you don’t receive unexpected notification emails when Nova scores candidates.
Ashby Integration: Automatic filtering for hired and archived candidates
Nova now automatically skips scoring for candidates in Ashby who have already been hired or archived. This saves you credits by only scoring candidates who are still being actively evaluated.Security: Easier sign in for existing customers
We added a “Sign in” button to the Nova homepage, making it simple for existing customers to access their accounts.Talent Pool: More accurate candidate fit assessments
Talent Pool searches now deliver fit ratings that better match recruiter expectations:- Overqualification detected: A candidate with 12+ years of experience for a role requiring 3-5 years is now rated Consider instead of Best Fit
- Location preferences respected: A US-based candidate is appropriately flagged when you’ve specified a preference for Latin America, rather than being rated Strong Fit
- Strong candidates not held back: Candidates who match your criteria but have small unknowns (like timezone not listed) are promoted to Best Fit instead of being penalized
- Missing info triggers follow-up, not rejection: If a profile doesn’t mention “startup experience” but doesn’t contradict it either, they’re rated Consider so you can ask during outreach—not automatically rejected
Calibration: Multi-file resume upload
You can now select and upload multiple resume files at once when running calibration, instead of adding them one by one. The upload shows progress as each file is processed.Talent Pool: Email notification when search completes
You’ll now receive an email when your talent pool search finishes, with a summary of candidates found (Best Fit, Strong Fit, Consider) and a direct link to view results. Feel free to close the page while searching - we’ll let you know when it’s ready.Talent Pool: See your team’s recent searches
The recent searches dropdown in Talent Pool now shows searches from all team members, not just your own. Each search shows who ran it (“by You” or the team member’s name) and when (“2 hours ago”), making it easy to find and reuse searches your colleagues have already run.Criteria Builder: Multi-select clarification questions
When building job criteria, you can now select multiple options for clarification questions instead of being limited to a single choice. This provides more flexibility when defining requirements that may have multiple valid answers.Criteria Builder: Improved scrolling and auto-save
The criteria builder now automatically saves your work when you click “Continue to edit” and scrolls to the top of each section for better visibility. This makes it easier to navigate through the criteria building process.Resume Upload: Encrypted PDF detection
The system now detects password-protected or encrypted PDF resumes before processing and provides clearer feedback when a resume cannot be processed due to encryption.Talent Pool Search: Give feedback on results
You can now provide feedback on talent pool search results to help improve future searches. The feedback interface shows clear visual indicators and improved messaging during the search process.
Calibration: Better handling of criteria changes
The calibration interface now better preserves scoring information across re-runs and provides clearer visual feedback when criteria have been removed. This prevents loading issues and improves the overall calibration experience.Talent Pool: Exclude already-applied candidates
You can now exclude candidates who have already applied to the job from Talent Pool search results. This helps you focus on net-new prospects and avoid reaching out to candidates already in your pipeline.Talent Pool: Preview rediscovery settings
You can now preview rediscovery gating settings before they take effect. This lets you understand which candidates will be eligible for rediscovery based on your configured rules without impacting live searches.Source Insights: Filter by score
Source Insights now supports filtering candidates by score ranges and organizes results into score buckets. This makes it easier to identify high-quality sourcing channels and understand score distribution across different sources.Settings: Re-invite team members
You can now send multiple team invitations to the same email address, making it easier to re-invite team members who may have missed or lost previous invitation emails.Analytics: Application source insights (beta)
You can now see detailed insights about where your job applications are coming from, helping you understand which recruiting channels are most effective and where to invest recruiting resources.Security: Passwordless two-factor authentication
You can now enable two-factor authentication without requiring a password, providing a more streamlined and modern login experience while maintaining security.Criteria Builder: Interactive clarification questions
The criteria builder now asks clarifying questions when job requirements are ambiguous or need more detail. This helps ensure the criteria accurately captures what you’re looking for in candidates.
Candidate Analytics: Indexed candidate chart
A new chart shows the share of indexed candidates over time, giving you visibility into how much of your candidate pool has been processed and is ready for search.Talent Search: Geographic proximity boosting
Candidates who are closer to your job’s location now rank slightly higher in search results, making it easier to find local talent without completely filtering out remote candidates.Talent Pool: Recent search history
You can now access your recent talent pool searches directly from the jobs table. A dropdown menu next to the search button shows the last 5 searches for each job, making it easy to revisit previous results without re-running searches.Talent Pool: Improved search relevance
Search results are now powered by an AI-driven query planner that better understands search intent and returns more relevant candidates. The system uses advanced ranking with intelligent query expansion to surface the best matches for your job criteria.Talent Pool: Configurable lookback date
You can now set a custom start date for syncing recent candidates into the talent pool from Settings. This allows you to control how far back in your ATS history to sync candidates, overriding the default 24-month lookback period when needed.Talent Pool: Qualitative fit scoring
Talent pool results now show qualitative fit assessments instead of numeric scores, making it easier to understand why a candidate matches your criteria and focus on the most relevant factors.Calibration: ATS candidate links
Candidates in the calibration interface now include direct links to their profiles in your ATS, making it easier to view full candidate details and take action without switching contexts.Calibration: Hiring manager feedback
Hiring managers can now provide structured feedback on scored candidates through a new calibration interface, allowing them to validate or challenge Nova’s recommendations and improve scoring accuracy.Calibration: Improved rerun comparison
After re-running calibration, the comparison view now makes it easier to see what changed and understand how adjustments to criteria affected candidate recommendations.Talent Pool: Streaming search results
Search results now appear progressively as they’re found and are automatically reranked for relevance, providing faster initial results and better ordering as the search completes.Talent Pool: Candidate location enrichment
Candidate profiles now include precise location coordinates and timezone information, making it easier to filter and understand candidate location details for distributed teams.Talent Pool: Improved results design
The search results table has been redesigned with clearer fit indicators, making it easier to scan and identify the best candidate matches at a glance.Talent Pool: Search from job row
You can now launch a talent pool search directly from the jobs table with a single click, making it faster to find candidates for any open role without navigating through multiple screens.Criteria Builder: Calibration setting
The criteria builder now includes an option to enable or disable calibration for each job, giving you more control over how criteria are refined based on candidate feedback.Talent Pool: Criteria matrix table
Talent pool results now display candidate qualifications in a matrix format, making it easier to compare how candidates stack up against each criterion at a glance.Application Review: Required screening questions
Applications now clearly highlight which screening questions are required, helping you quickly identify candidates who have completed all necessary responses.Interview Guides: Export PDF
You can now export interview guides as PDF documents, making it easy to print or share structured interview questions with hiring team members who prefer offline access.Criteria Builder: Importance levels
You can now set importance levels for each criterion when building job requirements, helping Nova better prioritize which qualifications matter most when scoring candidates.Job Management: Requisition ID field
Jobs now support a requisition ID field that syncs with your ATS, making it easier to track and reference positions across systems.Jobs: Automatic scoring when run limits change
When you adjust run limits for a job, Nova now automatically processes applications without requiring manual intervention, ensuring your job postings stay up-to-date with scoring results immediately.Jobs: Per-job application run limits
You can now set application processing limits per individual job through a new Job Settings dialog. The Output Language setting has also been moved into this dialog for easier access to all job-specific configurations.Jobs: Job ID column in jobs table
The jobs table now displays a Job ID column with tooltips, making it easier to reference and identify specific job postings when communicating with support or teammates.Workable Integration: Stage handling guidance
For Workable integrations, the UI now displays helpful notes about how certain hiring stages are handled, providing clarity on which candidates will be scored by Nova.Jobs: Fixed individual filter clearing
Clear filter buttons in the jobs table now properly reset individual filters instead of all filters at once, giving you more precise control over your job listing views.Nova Status: Last synced timestamp
Nova status tooltips now show when applications were last synchronized from your ATS, helping you understand data freshness at a glance.Analytics: Combined jobs and applications chart
The analytics dashboard now displays a unified chart showing both jobs and applications over time, making it easier to see trends and correlations in your hiring data.Workable Integration: Configurable rating scale
Workable users can now configure how Nova’s scores map to Workable’s star rating system, allowing you to customize the rating thresholds to match your hiring process.Billing: Monthly job creation chart
Added a new chart to the billing page showing your monthly job creation trends, helping you track hiring activity over time.Workable Integration: Stage filtering for candidates
Workable integration now supports filtering candidates by stage, giving you more control over which applicants are synced and scored.Integrations: Improved ATS setup experience
Enhanced the ATS integration setup interface with clearer instructions and better documentation, making it easier to connect your applicant tracking system.Criteria Builder: Custom brand specification
You can now specify preferred brands or competitor companies when building job criteria, helping the AI better understand the type of experience you’re looking for.Criteria Builder: Retry and skip brand research
Added the ability to retry or skip the brand research step when building criteria, giving you more control over the criteria creation process.Integrations: Smart auto-tagging
Introduced AI-powered automatic tagging that intelligently categorizes candidates in your ATS based on their qualifications and experience, saving time on manual organization. Available for TeamTailor and Greenhouse.Tags: Auto-generate AI instructions
New feature that automatically generates concise AI instructions for tag configurations, streamlining the setup of custom tagging rules.Integrations: Shorter shareable job links
Job links are now significantly shorter and easier to share, moving from lengthy URLs to a cleaner/l/{code} format for a more professional appearance.