// features · the full engine

The 19 autopilots working while you sleep.

Hey — I'm Pablo. I built LaneBolt from a cracked phone in Australia while caring for my parents full-time. That means every feature had to run on its own — no live sales team, no ops room, no support desk. Here's exactly what runs in the background so you can quote freight and go to sleep.

// jump to any autopilot
01 · 5-agent quoting

The Team Lead + four specialists that draft every freight quote.

Pricing, Route, Customs, Risk, and a Team Lead that stitches it all together into one defensible customer-facing quote.

You paste a shipment. Team Lead splits the job. Pricing hits live lane rates. Route picks ocean vs air vs rail based on urgency + margin. Customs surfaces the paperwork. Risk flags red-sea diversions, port congestion, or capacity tightness.

The whole orchestra runs sequentially in under 60 seconds. Every agent leaves a trace so you can defend the number when your customer says "but Flexport quoted less."

It's the single feature that killed my Google Sheets forever. Freight quoting used to eat 45 minutes per RFQ. Now it's the click before the coffee finishes brewing.

Every quote ships with a public share link — 3D route globe, benchmark, countdown timer, pay-in-crypto button. Your customer never asks "can I get this in a PDF?" again.

  • Multi-mode: ocean, air, rail, truck
  • Ship-ready output: line items, currency, incoterms, ETA, expiry countdown
  • Public share link with 3D route + native pay (Card/PayPal/USDC)
  • Full orchestration trace so you can defend the number line-by-line
demo preview
// demo · full 5-agent quote in 47 seconds
// live capture coming soon
See it live in your workspace
02 · ai ops manager

A chief-of-staff that reads every quote and tells you what to touch first.

Auto-triage on every new quote. P0/P1/P2 priority, reason, assignee, response deadline — all set before you even log in.

Every new quote passes through the Ops Manager the moment it lands. Claude reads the shipment, your team, and your open queue, then decides: this is a P0 today, this can wait until Thursday, this belongs to Sara because she owns the LATAM lane.

At 8am UTC it sends the team owner a warm 2-paragraph digest naming specific quotes and sequencing the day. No dashboards to scan. No Slack to check. Your morning coffee tells you what matters.

You can also ask it plain-English questions like "which quote is my top priority?" or "how many overdue this week?" and it answers with real numbers grounded in your team's data.

This runs hourly in the background. It'll silently backfill triage on anything you seeded through the API. It'll flag overdue quotes with an amber pill on your dashboard. It never sleeps.

  • Auto-triage on every new quote (P0/P1/P2 + reason + assignee)
  • Morning digest 8am UTC — warm, specific, sequenced
  • Natural-language Q&A over your quote queue
  • Team load balance chart so no one drowns quietly
demo preview
// demo · morning ops digest, real team
// live capture coming soon
See it live in your workspace
03 · customer god mode

Your customers see the money LaneBolt made them — with confetti.

Every forwarder gets a live pulse dashboard: savings vs. broker margin, win rate, top lane, team seats, plus milestone celebrations at every threshold.

The old freight software makes forwarders feel like they're subsidising an accounting department. LaneBolt does the opposite. Every customer sees their own Pulse: "$14,320 saved vs. a traditional 15% broker markup, 4 of 7 booked, top lane Shanghai→Rotterdam."

The moment they cross a threshold — first booking, first teammate, $1k saved, 100 quotes — a full-screen confetti overlay unlocks with a share-ready branded PNG. Their name. Their number. Their brand on top, LaneBolt in the footer. Free viral loop.

Ten milestone thresholds fire automatically as they grow (first_quote, first_booking, first_teammate, savings_100/500/1k/5k/10k, quotes_10, quotes_100). Every celebration includes a native share sheet so a happy customer can post to LinkedIn in one tap.

You (the LaneBolt owner) also get a live push notification when your customer hits a big one, so you know exactly when to reach out for a testimonial. Their win is your win.

  • Live pulse: savings vs 15% broker margin, win rate, top lane
  • 10 milestone thresholds, full-screen confetti celebrations
  • 1080×1080 branded share card auto-generated per milestone
  • Native share to X / LinkedIn / WhatsApp / iOS-Android share sheet
  • You get a push ping when a big win fires — testimonial-timing gold
demo preview
// demo · $1,000 saved celebration + share card
// live capture coming soon
See it live in your workspace
04 · reply autopilot

Claude answers customer emails in your voice while you're asleep.

Inbound customer reply hits your inbox. AI drafts a warm on-brand response. Sends automatically (or holds for you to approve).

Two entry points. Public webhook that accepts Resend Inbound / SendGrid Parse / Postmark shape. Or paste-a-reply from Gmail if you'd rather trigger it manually. Either way the autopilot core is the same.

It answers pricing questions ($199 / $499 / $1,499 with a 14-day trial and no card). It routes demo requests to the auto-demo on the landing page instead of a live call. It handles the "we already use Flexport" objection with the complement framing.

Opt-out safety is built in. If a customer says "unsubscribe" or "not interested" the AI silently adds them to a suppression list and the nurture sequence stops forever. No apology-loop, no shame.

There's a turn cap. After 3 back-and-forth replies the AI decides the lead is warm enough for you to take over and logs a handoff marker in the outbox. Never accidentally alienates a real buyer with an infinite bot.

  • Warm, on-brand replies in Pablo's voice
  • Opt-out safety: catches 12 phrases, updates suppression list forever
  • Turn cap: max 3 back-and-forths, then a human handoff flag
  • Auto-send OR draft-only modes per your comfort level
demo preview
// demo · sample reply thread, 3 AI turns
// live capture coming soon
See it live in your workspace
05 · freight intel weekly

A 700-word freight blog post + a LinkedIn post — every 7 days, forever.

You never write another piece of content. Claude ships a full markdown post + matching LinkedIn caption weekly, then emails every waitlist subscriber.

Seeded topic queue includes "Why the 45-Minute Freight Quote Is Dead," "5 Hidden Fees on Your Freight Invoice," "Every Incoterm 2026 Explained," and 9 more. When the queue runs dry, the AI invents 3 fresh topics from what's already been published.

Every post is 700-1000 words with a real thesis, 3-5 sub-sections, and a warm founder sign-off. Nothing generic. Nothing that reads like an ad.

The moment it publishes, every non-opted-out waitlist subscriber gets an email with a CTA button to the public post URL. Content marketing without the content-marketing team.

There's also a LinkedIn post drafter (`/linkedin`) with pre-tuned templates for founder-in-public updates, milestone celebrations, freight tips, and first-customer hunt posts. One tap → warm on-brand caption.

  • 700-1000 word blog + 120-180 word LinkedIn caption every 7 days
  • Emails every non-opted-out waitlist subscriber automatically
  • 12+ seeded topics; AI invents fresh ones when queue is empty
  • LinkedIn drafter with 6 vibe templates for on-demand posts
demo preview
// demo · read this week's post at /insights
// live capture coming soon
See it live in your workspace
06 · auto-hazard warnings

Claude scans every lane for real-world risk before the quote leaves your hands.

Red Sea diversion. Typhoon. Port congestion. Capacity tightness. All flagged automatically, with a suggested action per hazard.

The moment a quote is generated, we hand the origin + destination + mode to Claude and get back 0-3 hazards affecting that specific lane over the next 14 days. Each hazard has severity (low/medium/high), a title, a detail line, and a suggested action.

Results are cached for 24 hours per lane so the same quote pair doesn't re-hit the LLM. If a run fails the quote still ships — hazards are additive, never blocking.

Hazards appear in two places. The private quote detail page shows a warm amber warning panel with a rescan button, severity chips, and per-hazard suggested actions. The public share page shows the customer-friendly version: "Your forwarder ran an AI risk scan…" right after the 3D route globe.

Clean lanes get a jade reassurance panel: "lane check · all clear." No empty state ever, no doubt in the customer's mind.

  • Every quote scanned automatically for 0-3 lane hazards
  • 24-hour cache per lane, no duplicate LLM calls
  • Severity chips + suggested action per hazard
  • Customer-facing panel on the public share page
demo preview
// demo · Shanghai→Rotterdam, 3 hazards found
// live capture coming soon
See it live in your workspace
07 · testimonial autopilot

Every warm customer reply is auto-mined into a testimonial you can approve.

Someone writes back happy. Claude extracts a 40-220 char quote. Queued on `/testimonials`. One tap to publish live on the landing page.

Every inbound reply passes through a quality scorer. If it's unambiguously positive, references LaneBolt / AI / speed, and scores above 0.35, we extract a testimonial-worthy quote and file it as `pending`.

You get a warm alert. On `/testimonials` you can edit the quote, add the author's name / company / role, and approve → the live landing page instantly renders the new card. No CMS, no re-deploy.

Complaints, questions, and mixed feedback are filtered out. You never accidentally publish someone lukewarm.

The landing page section ("03.5 · what forwarders say") auto-hides when the approved list is empty. Zero fake testimonials, ever. First real one you approve is the first the world sees.

  • Quality-scored mining of every inbound reply
  • Pending / approved / rejected workflow with inline edit
  • Landing section auto-hides when empty (never a fake card)
  • Approved testimonials render instantly, no re-deploy
demo preview
// demo · approve a testimonial in one tap
// live capture coming soon
See it live in your workspace
08 · auto-dunning

Warm Day-1 / Day-3 / Day-7 emails when a card bounces — never threatening.

Stripe fires `invoice.payment_failed`. Claude drafts a warm, non-shaming email with a one-click update-card link. Repeats gently on Day 3 and Day 7.

The tone shifts subtly across the sequence. Day 1 is casual: "your card just didn't go through, happens all the time." Day 3 is a nudge: "the account may be limited soon." Day 7 is the last warm handshake before service pauses.

Every email tells the customer their data / quotes / team are safe and untouched. Card-update is one tap to `/billing`. If Stripe fires `invoice.payment_succeeded` in between, dunning auto-stops.

Full opt-out respect. If the customer previously said "stop emailing me," dunning still respects the suppression list because losing a customer over a bounced card twice is worse than losing them once.

Every send is logged to the Comms outbox as `dunning_day1/3/7` so you have full transparency in `/comms` if a customer ever calls you upset.

  • Day-1 / Day-3 / Day-7 warm cadence, tone shifts progressively
  • Never threatens — every email says "your data is safe"
  • Auto-stops when Stripe reports payment succeeded
  • Respects opt-out list, logs every send to `/comms`
demo preview
// demo · sample Day-1 dunning email
// live capture coming soon
See it live in your workspace
09 · sunday wrap-up

Every Sunday at 8pm — a two-line email that says "go rest, everything ran itself."

AI aggregates the week: emails sent, replies handled, DMs drafted, new signups, quotes created, revenue, and warm leads. Two-paragraph email lands on your phone Sunday night.

It doesn't feel like a report. It feels like a friend texting a status update. "Hey — quick Sunday wrap-up so you know your business ran itself this week. AI sent 47 emails, handled 12 customer replies, drafted 8 DMs, welcomed 5 new signups."

Warm leads (customers who wrote back multiple times in the last 14 days) are called out by email so you know exactly who to peek at Monday morning.

Zero KPI overload. Zero jargon. Zero dashboard. Just: "you did nothing. Everything's fine. Go rest."

Also on Sunday morning: the Wins Digest email. Real customer signups, real quote counts, milestones unlocked. Fires Saturday 22:00 UTC (= Sunday 8am your time in Australia). Idempotent per ISO week.

  • Sunday 20:00 UTC wrap-up + Saturday 22:00 UTC wins digest
  • Aggregates every autonomous action across the week
  • Names warm leads by email so you know who to nudge Monday
  • Idempotent — never sends twice for the same week
demo preview
// demo · sample Sunday wrap-up email
// live capture coming soon
See it live in your workspace
10 · monthly linkedin card

1st of every month — a branded 1080×1080 share card + AI-drafted caption in your inbox.

PIL renders a LaneBolt-branded PNG summarising the last 30 days. Claude drafts a LinkedIn caption in your voice. Email arrives with the PNG attached, ready to post.

Zero design work. Zero copywriting. First of the month, 08:00 UTC, an email lands with subject "Your monthly LaneBolt share card is ready (September 2026) 📣." The PNG is attached. The caption is in a highlight-quote block ready to paste.

The card renders the four stats that actually matter: real signups, real quotes, customer savings, milestones unlocked. Only shows non-zero numbers so a quiet month doesn't get a fake fluff-fest.

The caption is under 110 words to fit above LinkedIn's "see more" fold. It only mentions non-zero stats, uses hashtags #Freight #AI #FreightForwarding, and ends with a soft invite.

Same rhythm you'd get from a $2k/mo social-media agency. Zero cost. Never misses a first-of-month.

  • 1st of month 08:00 UTC, auto-emailed to every platform owner
  • 1080×1080 branded PNG (same generator as customer milestone cards)
  • Claude-drafted caption, <110 words, hashtag-optimised
  • Idempotent — never sends twice for the same month
demo preview
// demo · September share card + caption
// live capture coming soon
See it live in your workspace
11 · ai financial insights

Claude reads your revenue every day and writes 3 short next-step lines.

Every dashboard load quietly runs an AI accountant across your bookings + top lanes + stale invoices. Refreshed daily. Cached per team so cost stays negligible.

Numbers alone don't tell you what to do. So we hand Claude your booked total, MoM change, top lanes, mode breakdown, and stale bookings, and ask for 3 warm ≤25-word next-steps in the forwarder's own voice.

Example insight (real): "Shanghai→Sydney alone drove $12,480 — ocean is your highest-value mode right now, worth prioritizing." No filler. No jargon. Grounded in your data.

There's a customer view ("what to double down on") and an owner view ("which forwarder to court for a testimonial, cash-flow risk if any"). Both cached per day, both hidden entirely on cold-start so nobody sees a hollow panel.

Full explainer collapsible on both panels — a two-tap "how is this calculated?" so first-time users instantly understand what they're reading.

  • 3 warm on-brand next-step lines, refreshed daily
  • Customer tier speaks to the forwarder, owner tier speaks to Pablo
  • 24h cache per (team, tier) so LLM cost stays flat
  • Hidden entirely on cold-start — never a hollow shell
demo preview
// demo · your dashboard once you close the first quote
// live capture coming soon
See it live in your workspace
12 · ai voice-to-quote

Speak a lane into your phone — get a full quote back.

Tap the mic on /quote/new, say something like 'Shanghai to Sydney, ocean, 8 tons, 40 CBM, express' — Whisper transcribes, Claude extracts, the form pre-fills. 20 seconds, hands-free, mobile-first.

Built for the reality that most solo freight forwarders quote from a phone, often while doing 3 other things. Voice-to-Quote turns a 20-second spoken brief into a fully-structured freight-quote request in under 10 seconds.

The pipeline: MediaRecorder captures your voice → OpenAI Whisper transcribes it in your language → Claude Sonnet 4.6 extracts origin, destination, mode, weight (converts tons and pounds to kg automatically), volume (converts cubic feet and FCL sizes to CBM), urgency, incoterm, cargo type, and free-text notes.

Handles imperfect speech beautifully — half-sentences, background noise, mixed units. If it can't hear a clean lane, it says so warmly instead of guessing. No competitor offers this for solo forwarders.

  • 20-second max recording with live countdown timer
  • Auto unit conversion — tons, pounds, cubic feet, FCL sizes
  • Works fully offline-degradable — form still works by typing
  • Warm error messages when we can't hear a clean lane
demo preview
// tap the mic on /quote/new
// live capture coming soon
See it live in your workspace
13 · predictive booking score + objection handler

0-100 booking score + 3 paste-ready objection responses on every quote.

Every new quote comes with a Predictive Booking Score (colour-coded jade ≥70 · volt ≥40 · amber <40) and 3 anticipated buyer objections with winning responses you can copy verbatim.

The moment your 5-agent AI team drafts a quote, a scoring agent grades how likely it is to close — grounded in lane + mode + urgency + customer domain signals. Fresh business-domain email = +20. Air + high urgency = +15. Weekend generation = -10. No customer email = -25.

Alongside the score, Claude drafts three anticipated objections a buyer might raise — 'this transit is too slow', 'I got a lower rate from X', 'the fuel surcharge feels high' — plus a paste-ready winning response for each one, in your voice.

Owner-only view on /quote/:id. Public share pages never expose the score or objections — those stay your competitive edge.

  • Score visible instantly on every new quote
  • 3 anticipated objections + winning responses per quote
  • One-tap copy on each response for use in emails/DMs
  • Never exposed to customers on the public share page
demo preview
// visible on every quote at /quote/:id
// live capture coming soon
See it live in your workspace
14 · ai win/loss autopsy

Every lost quote gets a 3-line coach draft in 5 seconds.

Mark a quote as LOST and Claude drafts: (1) why we probably lost, (2) one sentence to say next time, (3) one adjacent lane pivot where you're likely stronger. Free per-quote coach on every miss.

Losing quotes silently is how solo forwarders stall — no signal, no learning, no adaptation. The Win/Loss Autopsy fixes that: every LOST quote triggers a warm 3-line post-mortem the moment you mark it.

Grounded in the lane, mode, total, predictive score, top anticipated objection, and any hint you type in (e.g. 'Flexport was 8% cheaper'). Claude returns concrete, non-waffly guidance — think of it as a friend coaching you after every miss instead of a spreadsheet judging you.

The autopsy is persisted so you can revisit it later, and surfaces on the Ops page so patterns become visible over time ('lost 3 quotes on Trans-Pacific ocean this month — worth pivoting?').

  • 3-line post-mortem: why · next time · pivot suggestion
  • Kicks in the moment you mark a quote as lost
  • Warm heuristic fallback if the LLM is unavailable
  • Persisted so you can trend the pattern over time
demo preview
// tap 'mark this quote as lost' on any open quote
// live capture coming soon
See it live in your workspace
15 · ai auto-follow-up cascade

Warm Day-2 + Day-5 follow-ups auto-sent to every unbooked customer.

Every quote with a customer email on your Autopilot plan auto-receives a warm Day-2 check-in and a Day-5 last-nudge — both drafted by Claude in your voice, both auto-sent while you focus on family.

The single biggest reason forwarders lose deals: they quote and forget. LaneBolt fixes this permanently. Every quote you generate with a customer email attached becomes eligible for the auto-cascade.

48 hours after generation, the Day-2 nudge fires: a gentle 'just checking the quote landed OK' with one concrete tweak offered (mode swap · consolidation · earlier pickup). 5 days later, the Day-5 last-nudge: 'quote expires in 2 days — anything I can adjust?'

Both messages are drafted by Claude with the lane, mode, total, and top anticipated objection baked in. Autopilot teams: auto-sent. Non-Autopilot teams: drafts visible in /comms for one-tap manual send.

  • Day-2 gentle check-in + Day-5 last-nudge
  • Fully drafted by Claude in your voice
  • Auto-sent for Autopilot teams, drafted for others
  • 3-day cooldown so quotes never get spammed
demo preview
// active on every quote with a customer email
// live capture coming soon
See it live in your workspace
16 · ai rate repricer

Proactive price-drop emails on every open quote when rates likely fell.

Daily 06:00 UTC scan of every unbooked quote 24h-7d old. Claude judges if the spot rate likely dropped 5-12% (fuel, capacity, season, geopolitics). If yes → drafts a 'great news — I can lower this to $X' email.

Freight rates move constantly. When your customer is comparing your quote to a competitor's fresher rate, you often lose without ever knowing. The Rate Repricer solves this: it checks daily whether the market has likely moved on every open quote, and pre-empts the competitor.

For each candidate quote, Claude analyses lane + mode + urgency + age against fuel volatility, capacity trends, seasonal shifts (Chinese New Year, hurricane season), and geopolitical events (Red Sea, Panama). Conservative bias — only recommends a re-quote when the drop is likely 5-12%.

Autopilot teams: auto-sends the drafted email to the customer with an updated total. Non-Autopilot: drafts land in /comms for one-tap review. 3-day per-quote cooldown so nobody gets multiple reprices in a week.

  • Daily 06:00 UTC across every open quote 24h-7d old
  • Only fires when drop is genuinely likely 5-12%
  • 3-day cooldown per quote prevents nagging
  • First tool of its kind for solo forwarders
demo preview
// runs automatically every morning on your quotes
// live capture coming soon
See it live in your workspace
17 · ai cold-prospect autopilot

5 hyper-personalised LinkedIn DMs drafted every Sunday.

Every Sunday 10:00 UTC LaneBolt analyses your top 3 booked lanes and drafts 5 warm, human-voice LinkedIn DMs to freight-forwarder prospects hand-picked to match your winning trades.

The other half of growth is outbound. Most solo forwarders skip cold outreach because it feels awkward and takes hours. LaneBolt turns it into a 90-second Sunday ritual: 5 fresh drafts land in your inbox before Monday coffee.

The AI picks one of 4 hooks per DM: city-focus ('noticed you run out of Perth'), lane-match ('you probably route Shanghai out — happy to send a fresh 60-sec quote'), boutique-fit ('LaneBolt sits at your size perfectly'), or founder-story ('I'm Pablo — solo caregiver-founder').

Each DM comes with a big blue 'Open LinkedIn profile' button and copy-paste-ready text. Approve or Skip in one tap. Prospects you skip are cooled-down for 30 days so they never re-appear stale.

  • Sunday 10:00 UTC — 5 fresh personalised DMs
  • One of 4 hooks picked per prospect (city / lane / boutique / founder)
  • Direct LinkedIn profile buttons in the email digest
  • 30-day cooldown per prospect after skip
demo preview
// hits your inbox Sunday morning · manage on /prospects
// live capture coming soon
See it live in your workspace
18 · ai customer sentiment radar

Save at-risk customers before they ghost.

Every morning 07:00 UTC LaneBolt scans every customer who's touched a quote in the last 90 days, reads their comms + quote history, and scores each HOT/WARM/COLD with a one-line reason and a specific next move.

The #1 hidden churn driver for freight forwarders: customers who go silent. They don't complain, they just fade. By the time you notice, they're already with a competitor.

The Sentiment Radar reads your comms_log (every autopilot email, every follow-up, every reply) + quote history (booked, lost, open, stale) per customer, and lets Claude judge the current sentiment. HOT = book-ready, replied recently. WARM = interested but quiet. COLD = silent > 14 days, unresponsive follow-ups.

The morning digest emails you the top 3 HOT customers (book-ready, spend energy here) + top 3 at-risk (cold or downgraded from warm) with a specific one-sentence next move for each. Zero-touch churn prevention while you focus on family.

  • Daily 07:00 UTC scan of every recent customer
  • Scores HOT / WARM / COLD with confidence + reason
  • Digest email lists top 3 hot + top 3 at-risk with next moves
  • Tier-change detection: notified when warm→cold happens
demo preview
// hits your inbox every morning · list at /admin/sentiment-radar
// live capture coming soon
See it live in your workspace
19 · ai case-study auto-generator

Every booked quote turns into a shareable LinkedIn card.

One tap on a booked quote produces a 4-part shareable card: headline · lane line · 3 concrete wins · pull-quote · warm CTA. Ready for LinkedIn, proposals, WhatsApp share.

Every win you close is invisible unless you tell someone. Most forwarders never share because writing a case study feels like work. The Case-Study Auto-Generator removes that friction entirely.

The AI reads the booked quote (lane, mode, weight, urgency, total) and drafts: a headline like 'Booked: Shanghai → Sydney in 60 seconds', a lane_line summary, 3 concrete wins ('locked-in rate before fuel surcharge shifted'), a shareable pull-quote in your founder voice, and a soft CTA ending with lanebolt.com.

Cards are cached per quote so re-opening is instant. A share button on QuoteDetail copies the pull-quote to your clipboard for one-tap paste to LinkedIn.

  • One-tap 4-part shareable card per booked quote
  • Includes founder-voice pull-quote ready for social
  • Cached forever per quote — never re-drafted
  • Powers your Sunday auto-post with fresh authentic content
demo preview
// visible on any booked quote at /quote/:id
// live capture coming soon
See it live in your workspace
// 19 autopilots · 0 humans required

Every one of these is already live in your workspace.

You don't turn them on. You don't configure them. You sign up, create your first quote, and every autopilot on this page starts running for you the same day.