ANTHIVA • USE CASE. RFQ Processing, Automated. From inbound email to ready-to-send quote — in minutes, not days.. 

Scene 1

From inbound email to ready-to-send quote — in minutes, not days.

Built on Anthiva  •  No-code AI workflows for the enterprise

What is an RFQ?. Setting the stage. Anthiva • RFQ Processing Automation. 

Scene 2

Request for Quote — the front door of every freight deal.

Shippers email carriers asking for pricing on specific shipments. Each RFQ contains lanes, weight, volume, equipment type and service level — but every shipper formats it differently.
 Speed and accuracy of the reply directly determine win rate.

From: ops@acme-shipping.com Subject: Pricing request — Q3 lanes
 Hi team,
 Need rates for:  • Chicago → Dallas, 24,000 lbs, dry van, weekly  • Atlanta → Miami, 18,500 lbs, reefer, 2x/wk  • LA → Phoenix, 12 pallets, LTL, ad-hoc
 Need transit times + fuel surcharge. Please reply by EOD Thursday.
 Thanks, Dana

The Problem. Why RFQ response is broken today. Anthiva • RFQ Processing Automation. 

Scene 3

Manual interpretation kills both speed and accuracy.

RFQs arrive as free-text emails, attached PDFs, spreadsheets, or even scanned screenshots — no two look alike.

Sales reps re-read each RFQ to identify lanes, weights, volumes, and service requirements line by line.

By the time a quote is built, the customer has already opened a competing carrier's reply. Win-rate suffers.

⚠  Industry data:  20–40 minutes per RFQ  •  30%+ of quotes never get sent in time  •  Win-rate drops with every hour of delay

Anthiva's Answer. An end-to-end AI workflow — no code required. 

Scene 4

An end-to-end AI workflow — no code required

From unstructured email → structured quote → sent or reviewed.

Parse RFQ emails into structured fields — lanes, weights, equipment, service level.

Standardize inputs and feed them into your existing pricing engine or rate sheet.

Auto-build comparable HTML quotes ready to drop into a reply.

Confidence-gate every quote — auto-send the easy ones, escalate the rest.

✓  Plugs into the tools you already use

Gmail  •  Google Sheets  •  Google Gemini 2.5 Flash  •  Any CRM via Composio  •  Slack, Teams, Salesforce, HubSpot — zero new vendor onboarding.

The Workflow at a Glance. Six steps from inbound email to logged quote. 

Scene 5

Six steps from inbound email to logged quote

New RFQ Email — Trigger. How the workflow gets started. 

Scene 6

Watches the configured Gmail inbox for incoming RFQ emails. The moment a matching email arrives — for example, subject contains 'RFQ' or comes from a known shipper domain — the workflow fires.

Attachment binaries forwarded to next step

Extract RFQ + Lanes — AI Agent. Turn unstructured email into structured data. 

Scene 7

Turn unstructured email into structured data

A Gemini 2.5 Flash agent reads the email, the body, and any attachments. It identifies each freight lane, its weight or volume, equipment type, service level and deadlines — and returns clean JSON.

Array of lanes with origin, destination, weight

Equipment type (van / reefer / flatbed / LTL)

Lookup Rates — Pricing Engine. Match each lane to your live rate sheet. 

Scene 8

For every extracted lane, query your Google Sheets rate book (or any pricing engine via Composio). Pull the base rate, fuel surcharge, accessorials and expected transit time.

Optional: customer-specific contract rates

Build Quote HTML — AI Agent. Compose the reply the rep would have written. 

Scene 9

Compose the reply the rep would have written

A second Gemini agent assembles a professional HTML quote. It mirrors your house style, includes a per-lane table, totals, validity window, and any standard terms your team already uses.

Auto-send Gate — Confidence Branch. Decide: ship it now, or send to a human?. 

Scene 10

🧭   Auto-send Gate — Confidence Branch

A configurable branch evaluates the quote's overall confidence and rule-based guardrails — value above $X, new customer, lane outside footprint — and routes the quote to one of two paths.

Branch decision: AUTO-SEND or HUMAN-REVIEW

Send or Draft — and Always Log. The two outcomes of the gate. 

Scene 11

Replies in-thread from your Gmail account

Customer receives quote in minutes — not days

When confidence < threshold or rule trips

Saved as a Gmail draft in the rep's mailbox

AI did 95% of the work — human owns the call

Reason for escalation captured for tuning

Confidence Gate, Explained. Your rules, your thresholds. 

Scene 12

auto_send IF     quote.confidence >= 0.85     AND quote.total_value < 25000     AND customer.tier IN ('A','B')     AND every(lane.origin IN our_footprint)
 ELSE     route_to_human     reason := first_failed_clause()

End-to-end example. From inbox to sent quote in 2 minutes. 

Scene 13

3 lanes • CHI→DAL, ATL→MIA, LA→PHX

Business Impact. What customers see in the first 90 days. 

Scene 14

Figures reflect customer benchmarks across logistics pilots. Actual impact varies by RFQ volume and pricing complexity.

Why Anthiva. Not a script. Not a chat box.. Anthiva • RFQ Processing Automation. 

Scene 15

Ops authors workflows. Engineers don't become the bottleneck. Every change is versioned and reviewable.

Gmail, Google Sheets, Salesforce, HubSpot, Slack, Teams. Connectors via Composio — zero new credentials to manage.

Every action logged. Confidence gates keep humans in the loop. PII handling, retries and rollback are first-class.

Frontier models with tool calling. Memory, structured outputs, evals built in — so quotes stay accurate as your business changes.

rfq_processing

Scene 1 (0s)

Scene 2 (13s)

Scene 3 (49s)

Scene 4 (1m 23s)

Scene 5 (1m 57s)

Scene 6 (2m 28s)

Scene 7 (3m 0s)

Scene 8 (3m 35s)

Scene 9 (4m 7s)

Scene 10 (4m 39s)

Scene 11 (5m 10s)

Scene 12 (5m 43s)

Scene 13 (6m 9s)

Scene 14 (6m 39s)

Scene 15 (7m 5s)

Scene 16 (7m 38s)

Scene 17 (8m 1s)