Everything, answered.

An operating system for executive search consultants and recruiters: 30 tools covering the complete search lifecycle — business development, client briefing, role decoding, sourcing, candidate evaluation, client presentation, interview management, offer closing, and post-placement care. The tools were built inside a working search practice and encode the disciplines of retained search — and every output hands you the next step.

Executive search consultants, recruiters, and talent teams. It's most powerful for retained-style work (senior, hard-to-fill roles) but every tool works for contingent recruiting too.

HunterNodes is a product of Axentia Global Private Limited, a company registered in New Delhi, India, with executive search operations across Asia. It was built inside a working search practice — the tools encode how consultants actually win and run mandates, not how software companies imagine they do. Contact: rs@axentiaglobal.com.

Yes — a free account starts your 7-day trial. Sign in with Google, LinkedIn, Microsoft, or your email; no card is needed. The account only identifies you and your licence — your mandates, candidates, and API keys still live in your own browser storage, never on our servers.

The clock starts the first time you open the app. You get the complete product for 7 days — all 30 tools, full output, nothing held back. When it ends, tool runs pause until you buy Pro, but nothing is deleted: every mandate, candidate, and document stays in your storage and you can still open and export all of it.

Open any tool and click '✨ Load example'. Every tool has realistic sample data (a VP Engineering search at a fictional robotics company) so you can run it immediately and judge the output.

Open 'The Playbook' in the left menu. It lists the nine stages in order, the tool for each step, how often to run it, and the rule that separates great work from average. If you're starting a real search: create a Mandate first, then run the Role Decoder.

One open search — the client, the role, the job description, comp, and notes. Create it once and every tool pre-fills from it automatically, so you never paste the same JD twice. All outputs attach to the mandate's history.

HunterNodes is the operating system; the AI engine is yours. You connect your own account from Claude (Anthropic), Perplexity, or OpenAI in Settings, and every run bills your account directly — typically 1–5 cents per run. That keeps your AI usage under your own control and billed at cost — you pay us for the operating system, not for tokens.

Claude: console.anthropic.com → API Keys → Create Key (starts with sk-ant-). Perplexity: perplexity.ai → Settings → API (starts with pplx-). OpenAI: platform.openai.com/api-keys (starts with sk-). All three need a payment method or small credit balance on their side — $5 goes a very long way.

Claude gives the best reasoning quality (scorecards, offer strategy) and supports live web search. Perplexity is excellent value and built for live search — great for the BD tools. OpenAI works for all writing tools but has no live web search here, so BD intelligence tools (Market Radar, Decision-Maker Finder) fall back to framework guidance instead of live findings.

Your key is stored only in your browser's local storage — never in any database. When you run a tool it's sent over HTTPS with that single request, used once to call your AI provider, and never logged or saved server-side. You can remove it anytime in Settings, and you can revoke the key itself at your provider whenever you want.

Yes — Settings has an optional Model field. Leave it blank for sensible defaults (claude-sonnet-5, sonar-pro, gpt-4o) or enter any model id your provider account has access to.

Click 'Test key' in Settings — it makes one tiny call (a fraction of a cent) and tells you clearly whether the provider accepted it.

If the site owner has configured a house key, runs use that. Otherwise tools show a clear message pointing you to Settings. Nothing breaks — you just can't generate output until an engine exists.

Yours. A key you add in Settings always takes priority, so your runs always bill your own account.

Start with a 7-day free trial — all 30 tools, your own AI key, no card required. After that, Pro is S$5 per seat per month, or S$50 per year (two months free) — less than a coffee a month to run your whole desk. Full details on the Pricing page.

Pro is paid through PayU's secure checkout — international cards accepted. Prices are shown in SGD; the charge is processed in the INR equivalent (shown at checkout) and your bank converts at its card rate. Your card details go to PayU directly and never touch our servers. After payment you'll hear from us within one business day to activate your licence.

Yes — invoicing is included with Pro. Reply to your activation email with your company details (GST details for Indian firms) and we'll issue it.

We email you a licence code, normally within one business day of payment. Open Settings → Pro licence, paste the code, and click Activate. The trial limit lifts immediately on that browser. Codes are tied to your email and expiry date, so please don't share them.

No. There is no auto-renewal and we do not store your card — we literally cannot charge you again. Your licence simply runs to its end date. We'll email you beforehand, and renewing is entirely your choice.

Because everyone gets a full 7-day free trial before paying, licence purchases are non-refundable once the code is issued. We do refund payments made in error, duplicate charges, or a payment where no licence was ever delivered — email us within 7 days with your PayU transaction ID. See the Refund & Cancellation Policy for detail.

Tool runs pause until you renew — but nothing is held hostage. Every mandate, candidate, and document you created stays exactly where it was (your browser, your folder, or your own database), and you can still open and export all of it. Renew and you're running again instantly.

Rough guide at current provider prices: a Role Decoder or Scorecard run costs about 1–3 cents on Perplexity or Claude Sonnet; long documents (Offer Playbook, Reference Kit) up to ~5 cents. A full search using 20 tool runs costs well under $1. Live-search BD tools cost the same order of magnitude.

A fair-use limit of 60 runs per hour per user protects the service from abuse. You'll see a friendly message if you hit it; it resets within the hour.

You choose, in Settings → Your data home. Default: this browser. Option 2: real files in a folder on your computer (pick a Dropbox/Drive-synced folder and cloud backup comes free). Option 3: your own free Supabase database for multi-device sync and team sharing. In every case, HunterNodes servers hold nothing.

Only the text needed to run a tool: your inputs (plus mandate context) travel to the AI provider you chose, via our server which relays them without storing anything. That's it. If you buy Pro, your payment details go to PayU directly — never through us.

Never. PayU handles the entire payment on their own secure page. We receive only your name, email, phone, the amount, and a transaction ID — enough to issue your licence and a GST invoice. We don't store your card and can't charge you again.

No advertising cookies, no analytics trackers, no profiling of any kind. HunterNodes stores things in your browser's local storage to make the product work (your data, preferences, AI key, trial date, licence). If you choose to sign in, one essential session cookie keeps you signed in. That's the complete list.

Yes — if you connect your own database (Settings → Your data home → option 3). Enter the same credentials on any device and you see the same mandates and pipeline. Browser and folder modes are single-device by nature (though a cloud-synced folder is readable on your other machines).

The Settings section where you choose where your work lives: this browser (default, zero setup), a folder on your computer (real files, covered by your own backups), or your own cloud database (sync + team workspace). Switching copies your data across, and a universal backup file works everywhere.

One teammate creates a free Supabase project (5-minute guided setup in Settings → Your data home), then shares the project URL + key with the team. Everyone who enters those credentials sees the same mandates, candidates, and outputs. Treat the credential pair like a team password.

Clearing your browser's site data deletes it, and private/incognito windows don't persist anything. Back up important outputs with the Word download button. Regular browsing, restarts, and updates are fine — localStorage survives those.

CV text you paste is sent to your chosen AI provider for that run — check your provider's data policy (both Anthropic and OpenAI offer no-training API defaults). Nothing is retained by HunterNodes. As always, handle candidate data per your local privacy laws (PDPA/GDPR).

They chain tools together. When a run finishes, the buttons under the output carry your inputs AND the output into the next logical tool. If the next tool has everything it needs, it runs automatically; if it needs something only you know (a CV, meeting notes), it pre-fills what it can and waits.

The one-line discipline rule the best firms follow for that step — e.g. 'Never send a CV you haven't scored.' They're not decorative: follow them and your work is retained-quality by construction.

Yes — every large input box accepts PDF, Word (.docx), and text files three ways: the Upload button, drag-and-drop onto the box, or pasting a file from your clipboard. Parsing happens in your browser.

A scanned PDF is a photo of text — there's nothing to extract without OCR. The tool detects this and tells you. Ask for a text version, or copy-paste manually.

The 1997-era .doc format isn't supported in-browser. Open it in Word, save as .docx or PDF, and drop that in.

No — every tool auto-saves your inputs as a draft in your browser and restores them when you return. 'Clear' wipes the draft.

Yes — the Stop button halts generation and keeps whatever has streamed. ⌘/Ctrl+Enter runs the tool from anywhere in the form.

Every output has Copy (clean markdown) and Word (.docx download) buttons. Search strings and alert queries appear in code blocks with their own one-click Copy buttons.

Every completed run saves to history automatically — see 'Recent outputs' on the dashboard, or the full history on each mandate's page. The most recent 300 outputs are kept.

A kanban board of every candidate in a search across 8 pipeline stages (identified → placed). Add candidates by name, move them with the arrows, and cards turn red after 7 days without movement — stalls become impossible to ignore.

The methodology prompts live in editable files (src/lib/prompts/ in the codebase) — the site owner can tune any tool's approach without touching code logic. Tell us what feels off and it's usually a 5-minute change.

Market Radar, Decision-Maker Finder, BD Company Finder, and Pitch Preparation — on Claude or Perplexity engines. They return real findings with source links.

It live-searches LinkedIn Jobs, Indeed, Glassdoor, and company career pages for your roles/location/industry, restricted to your chosen window (past 24 hours, week, or month). Every posting comes with its direct link, and companies with multiple openings are flagged as retained-search signals.

Two honest reasons: (1) narrow filters over a short window genuinely yield few senior postings — that's reality, not a bug; (2) live search sees what's publicly indexed, and some job boards hide listings from crawlers. The tool says so and suggests widening. For guaranteed coverage, use the alert stack Market Radar sets up on its first run — those alerts fire inside your own LinkedIn/Google accounts where nothing is hidden.

They are explicitly instructed never to — live tools may only report items actually found in search, each with a link, and must say 'not publicly findable' rather than guess. That said, always click the link before acting on anything — that's your professional duty and takes 5 seconds.

No scraping of logged-in LinkedIn happens. Live tools use your AI provider's web search over publicly indexed pages — the same pages a Google search sees. Market Radar's alert stack goes further by having YOU set up alerts inside your own LinkedIn/Google accounts, which is fully within their terms.

AI models are not deterministic, and the live web changes by the hour. Structure stays consistent (that's the methodology); wording and specific findings vary. For live tools, later runs may genuinely find newer postings.

1) Settings → Test key (expired key or empty provider balance is the #1 cause). 2) If the error mentions a model, your account may not have access to it — clear the Model field to use the default. 3) Rate limit message: wait a few minutes. 4) Very long inputs: trim to under ~60,000 characters.

Either Stop was pressed, the connection dropped (rerun — your inputs are drafted), or the response hit the length ceiling (rare; split the task).

Selecting a mandate overwrites mapped fields (like the JD) with the mandate's values — that's the pre-fill working. Switch back to 'No mandate' and click Load example again.

Current Chrome, Edge, Arc, Safari, and Firefox on desktop. It's responsive on mobile, though consultants mostly drive it from a laptop alongside LinkedIn.

Tell the Axentia team — the product iterates weekly and real usage feedback drives the roadmap.

Team accounts with cross-device sync and shared mandates, per-mandate Playbook progress checklists, Excel exports for lists and grids, calendar (.ics) export from the Process Calendar, and a Chrome extension so tools are available beside LinkedIn.

Yes — everyone opens the site, adds their own AI key in Settings, and works independently. Their data stays on their own machines until team accounts arrive.