Settings

Settings (/settings) is the configuration surface, split into sub-tabs: Job Preferences (keywords, locations, sources, exclude words), Subscription (tier, billing, plan switch), Profile (name, email, theme, AI provider, resume, career status), API Keys (automation tokens), and AI Keys (your own AI provider credentials).

TL;DR — New here? Set Profile (career status + target level) first, then Job Preferences (keywords + at least one location) to start the matcher. Each tab has its own ?tab= URL.

Quick start — first-run setup

If you've just signed up, the order to fill these in is:

  1. Profile — set name, career status, theme. Career status drives the recommended cadences shown on Insights.
  2. Job Preferences → Keywords — add 3 keywords (Free) or up to the tier cap (Pro). The matcher uses these to filter postings.
  3. Job Preferences → Locations — add at least one location (or "remote" — captured as a location string).
  4. Job Preferences → Sources — toggle on the boards you want the matcher to pull from. Free is capped at 4 public sources; Pro unlocks all.
  5. AI Keys tab — optional. AI is already included in your plan; add your own key only if you want unmetered usage that doesn't count against your monthly allowance. Saved keys are only used while the Use my own API keys switch is on — leave it off to stay on platform-funded AI.

After Profile + Preferences, the next ingest cycle (5am or 2pm Pacific) will start populating your Job Search inbox.

Job Preferences tab

Keywords

Each keyword is matched against job titles + descriptions. Multi- word phrases are exact-match ("vp engineering" matches that exact phrase). Stems are auto-handled — engineer matches engineering, engineers.

Tier caps:

  • Free: 3 keywords
  • Managerial: 10
  • Leadership: 25
  • Executive: unlimited

The platform de-dupes job postings before they reach you, so a job matching multiple keywords only shows up once.

Locations

Each location is matched against the job's location + remote-type fields. Use:

  • City names (San Francisco, CA)
  • States (CA, California)
  • Countries (United States, UK)
  • The literal string remote to match remote-only roles

Tier caps:

  • Free: 1 location
  • Managerial: 3
  • Leadership: 5
  • Executive: unlimited

A job is shown if it matches any location — locations are OR'd, not AND'd.

Sources

Toggle which job boards the matcher pulls from. Each source has a type — public (no auth), API-key (needs your key), or paid (scraped via Apify, platform-only).

Free tier: up to 4 public sources active at once. Pro: all sources.

If a source needs your own API key (Adzuna, Reed, etc.), add it under the AI Keys tab before toggling the source on. Without the key, that source returns nothing.

Exclude words

Drop words here that you never want to see in matched jobs (e.g. junior, intern, a specific company you're avoiding). Any posting whose title or description contains an excluded word is filtered before reaching your inbox.

Exclude words are case-insensitive. Use whole words; partial matches aren't supported.

Subscription tab

Shows your current tier (Free / Managerial / Leadership / Executive), the next renewal date, and links to:

  • Switch plan — upgrade or downgrade via Stripe
  • Manage billing — Stripe customer portal (update card, change cycle, cancel)
  • Apply discount code — if you have one

Plan changes take effect immediately; pro-rata billing is handled by Stripe on the next invoice.

During the invitation-only pilot, this tab shows a Pilot access card instead of upgrade/billing buttons — your tier is free until the date shown. Contact us if you need more time.

For deeper detail on what each tier unlocks, see Subscription.

Profile tab

The Profile tab has several sub-sections:

  • Basics — first name, last name, email (read-only — change via account flow)
  • Career status — drives the recommended cadences on Insights. Options: Actively looking / Open to opportunities / Building skills (employed) / Market awareness only / Not active
  • Theme — Light / Dark. Persisted server-side so your preference follows you across browsers + devices.
  • Timezone — your IANA timezone, auto-detected from your browser the first time you sign in and overridable here. It drives the calendar boundaries the app calculates in your local day rather than the server's: goal periods (daily / weekly / monthly resets), your monthly AI quota window, and report month/quarter labels. Clear the field to fall back to auto-detect.
  • AI provider — pick your default AI provider (Claude / Gemini / ChatGPT / OpenRouter / DeepSeek / Mistral / Grok / Groq / Together / Fireworks). Used everywhere in the app that has AI features (interview answers, application Q&A, role tips, etc.). The exact model identifier within each provider is admin-configured under /admin/ai → AI Setup → Platform-funded AI providers. When an administrator enables a self-hosted endpoint, a Local (private) option also appears — it keeps every prompt on WaypointCareer's own infrastructure with no third-party egress, and is metered against your monthly allowance just like the cloud providers.
  • Resume — upload your master resume (PDF). The AI uses the text for grounding answers + tips. One per account; uploading a new one replaces the previous.

AI Keys tab

Your own AI provider credentials (formerly under Profile). AI is funded by the platform and included in your plan, so a key isn't required. Add your own key for a provider only for unmetered usage on that provider, billed to your own account and not counted against your monthly allowance. A single Use my own API keys switch gates whether saved keys are used at all — while it's off, platform-funded AI is used even if you've stored a key. (This tab is separate from API Keys, which issues wcsk_… automation tokens for the public API.)

When your own key is required

On some accounts the switch is locked on and the tab shows an "Add your own AI provider key" notice. This applies to accounts that aren't currently paying for a subscription — including promotional and invite-based signups that were granted a tier for free, and checkouts that came to $0 through a discount code.

On those accounts AI runs entirely on your own provider key: nothing is drawn from the included monthly allowance, and nothing is metered here. Until you add a key, AI buttons stay disabled and explain why. Several providers have a usable free tier, so this usually costs nothing. Once a paid subscription is active the switch unlocks and the included allowance applies again.

Which provider to add

If you're adding a key only to satisfy the requirement above, OpenRouter is usually the cheapest and simplest choice: one key reaches 300+ models from many vendors (OpenAI, Anthropic, Google, Meta, Mistral, DeepSeek and others), and you pick which one to use from the Browse list next to the Text model field. Cheap models run well under $0.10 per million tokens, and some are free at reduced rate limits. Keys start with sk-or- and come from openrouter.ai/keys.

Two things to know about OpenRouter specifically:

  • Model ids include the vendor prefix — openai/gpt-4o-mini, not gpt-4o-mini. Picking from Browse gets this right for you.
  • It's used for text only. Draft images in Voice Composer still come from your OpenAI or Gemini key, so add one of those too if you want unmetered image generation.

Rate limits on OpenRouter depend on the model you choose rather than on OpenRouter itself, so a free model variant (ids ending in :free) can be rate-limited even with credit on your account.

If you would rather go direct to a vendor than through a router, three others are available, and all three are cheaper than the big US labs:

  • DeepSeek — the cheapest option overall. Its default model bills at $0.14 per million input tokens and $0.28 per million output, and repeated prompts get a large cache discount on top. Keys start with sk- and come from platform.deepseek.com. Worth knowing: requests are served from infrastructure in China, and DeepSeek has announced peak-hour pricing at double the standard rate.
  • Mistral — an EU provider, so your prompts stay in Europe. Its default (Mistral Large 3) is $0.50 in / $1.50 out per million tokens. Keys come from console.mistral.ai; a new account has to be activated for paid usage before the API answers.
  • Grok (xAI) — $2 in / $6 out per million tokens on grok-4.5. Keys start with xai- and come from console.x.ai.

Three more providers host open-weight models (Llama, Qwen, MiniMax, Kimi and similar) rather than publishing models of their own. They are worth knowing about for speed and price:

  • Groq — by far the fastest. It runs models on its own LPU hardware, so replies arrive in a fraction of the time other providers take. Its default bills at $0.59 in / $0.79 out per million tokens, and there is a free tier that is genuinely usable — capped per day rather than per minute. Keys start with gsk_ and come from console.groq.com.
  • Together AI — 100+ open models on serverless endpoints; the default is $0.30 in / $1.20 out per million tokens. Keys come from api.together.ai.
  • Fireworks AI — a strong lineup of very large reasoning models; the default is $0.95 in / $4.00 out per million tokens. Keys come from app.fireworks.ai.

For all three, model ids carry a prefix — MiniMaxAI/MiniMax-M3 on Together, and a full path like accounts/fireworks/models/kimi-k2p6 on Fireworks — so pick from Browse rather than typing one in. On Fireworks, only serverless models can be used with a plain API key; Browse lists exactly those.

Like OpenRouter, every one of these is text only — Voice Composer draft images still need an OpenAI or Gemini key.

Choosing your own models

Each AI provider also offers two optional model fields:

  • Text model — used for tips, prep, answers, and drafts.
  • Image model — used for Voice Composer draft images. Claude has no image field, because Anthropic does not generate images.

The popup opens with the fields first — API Key, Text model, Image model — and the provider's notes, limits, and signup link underneath in small print.

Model ids are not secrets, so unlike your API key each field opens prefilled with the model currently in force, and the label repeats it as in use: …. Clear a field and save to drop the override and go back to the platform default.

Browsing the available models

Next to each model field is a Browse button. It lists the models your own key can actually see on that provider, sorted into Text and Image, and fills the field with whichever one you pick.

  • It needs a saved key — a key typed into the form above hasn't reached the server yet, and the list is fetched with your key. Until one is saved the button stays disabled and says so.
  • Which models appear depends on your provider account's plan or tier. If one you expect is missing, enable or upgrade it in the provider's own console first, then reopen the list.
  • Text vs Image is a best-effort guess — providers don't reliably publish which models generate images. Turn on Show all models if something looks like it's filed under the wrong type.

Two more things to know:

  • Model fields apply only while "Use my own API keys" is on. A model override has no effect on a platform-funded call, so the fields are disabled when the switch is off.
  • The model id is checked against the provider when you save it. A typo or a retired identifier is rejected with the provider's own wording rather than failing later, mid-generation. If the provider can't be reached, the value is saved with a warning instead of being blocked — an outage shouldn't stop you fixing your configuration.

Notifications tab

Controls the WaypointCareer digest — the short weekly or monthly summary of your progress, pending reading, and any platform announcements. Three options:

  • Weekly (default for new accounts) — a digest every 7 days.
  • Monthly — every 30 days. Same content, less frequent.
  • Off — no digest. Welcome emails, password resets, and account notices still arrive — those are transactional and aren't gated by this preference.

The tab also shows when your last digest was sent so you know whether the next one is due.

Every digest email also carries a one-click Unsubscribe link in the footer (RFC 8058 — your mail client's native "unsubscribe" button calls it too). Clicking it flips this preference to Off.

See the full guide: Email & Notifications and Weekly / Monthly digest.

Tips

  • Career status matters. It drives the recommended cadences on Insights, default goal targets, and which goal templates surface as "suggested" on first run.
  • AI is included — top up if you run out. Every plan has a monthly AI allowance; Free's is small. If you hit it before the month resets, buy a usage credit pack from Settings → Subscription (credits roll over and never expire), or add your own provider key for unmetered usage on your own account.
  • Resume changes ripple slowly. The AI features that ground in your resume (interview answers, application Q&A) re-read the resume on each generation, so a new resume is reflected immediately for new generations — but old generated content is unchanged.

Related