Skip to content
Book a demo

InsightMesh · Intelligent Enterprise Query

Plain-English questions over your operational data, your databases and your documents — with specialist agents that turn supply-chain data into optimization recommendations.

  • Powers a Retail & FMCG supply chain command center in Microsoft Teams
Book an InsightMesh demoSee it in action

Proven results

  • Reduced logistics costs through intelligent consolidation
  • Improved pallet utilization
  • Faster operational decisions
  • Natural-language access to supply chain insights, inside Microsoft Teams

Read the case study

The InsightMesh family

Each assistant answers in plain English. Pick the one that matches where your answers live.

  • Logistics copilot

    Five specialist agents for supply-chain insight, order consolidation, pallet utilisation and drop-point centralisation.

  • BI agent

    Plain-English questions turned into SQL over your dataset, with tables, charts and a Word report.

  • Document chat

    Questions over PDF, Word, Excel, PowerPoint, audio, images and web pages, with page-level citations.

  • Unified data access (Roadmap)

    One conversational interface across structured and unstructured sources.

Product walkthrough

A simplified run-through of the InsightMesh workflow, step by step.

InsightMesh

Illustrative demo

Sample data · not a client resultA plain-English supply-chain question, routed to a specialist agent, computed by deterministic algorithms and explained step by step.Illustrative InsightMesh demo with sample data. A user types "Can we cut transport cost by consolidating next month’s deliveries to our top 5 customers?". Four steps run in order: parse intent, route to specialist (a supervisor agent hands the question to the Dynamic Order Consolidation agent), optimize, and explain. A routing map highlights the path from top 5 customers, next month and transport cost to the consolidation agent, then through its truck-loading and costing algorithms to the shipments, rate card and SKU master data. In the optimize step, shipment windows are tested, trucks are re-loaded and the rate card is applied. The answer: a 3-day shipment window cuts deliveries from 5 to 3 a week and raises pallets per order from 4.2 to 6.8, lowering projected transport cost per pallet. A before-and-after chart shows the same figures. The result is marked as computed by the Dynamic Order Consolidation agent from the rate card and shipment history, and tagged "Generated without BI development".

How it works

  1. 01 You

    Ask

    Type a business question in plain English, in the web app or in Microsoft Teams.

  2. 02 Supervisor agent

    Understand

    Reads the intent, rewrites follow-ups as standalone questions and keeps the conversation on topic.

  3. 03 Supervisor agent

    Route

    Hands the question to the right specialist — insights, consolidation, pallets or drop points.

  4. 04 Specialist agent

    Compute

    Deterministic algorithms do the maths — truck loading, rate-card costing, shipment windows.

  5. 05 Specialist agent

    Explain

    Returns the answer with a chart and a before-and-after table of cost, pallets and delivery days.

  6. 06 Supervisor agent

    Follow up

    Suggests the next questions worth asking, and remembers the context when you ask them.

Capabilities

Ask your operational data, databases and documents questions in plain English — with specialist agents that turn supply-chain data into optimization recommendations.

  • 01 Natural Language Analytics

    Ask business questions in plain English and get an answer, a chart and suggested next questions.

  • 02 Databases & Documents

    Query structured data through plain English converted to SQL, and chat with documents — PDF, Word, Excel, PowerPoint, audio, images and web pages — with page-level citations.

  • 03 Decision Intelligence

    Insights, trends and optimization recommendations, with before-and-after comparisons of cost, pallets and delivery days.

  • 04 Supply-Chain Optimization Agents

    Specialist agents for order consolidation, pallet utilisation and drop-point centralisation, built on deterministic truck-loading and rate-card costing.

  • 05 Conversational Memory

    Follow-up questions keep their context, and a guardrail keeps each assistant on its topic.

  • 06 Web and Microsoft Teams

    Use InsightMesh in a web app with interactive charts, or inside Microsoft Teams.

Use cases

  • Cut transport cost by consolidating deliveries

    Customers get several part-loads a week, and nobody has time to model a better delivery pattern.

    1. Ask
    2. Route
    3. Compute
    4. Explain

    Shipment windows and delivery-day schedules are tested against the rate card, with projected savings side by side.

  • Fill pallets better

    SKUs ship on half-empty pallets, and storage and handling costs add up.

    1. Ask
    2. Route
    3. Compute
    4. Explain

    Layer patterns, pallet heights and double stacking are tested per SKU, showing the effect on pallet count and cost.

  • Answer questions from your documents

    Policies, contracts and reports hold the answer, but nobody can find the right page.

    1. Upload
    2. Ask
    3. Answer
    4. Cite

    Answers come back with the page they came from, highlighted — with translation and read-aloud when you need it.

Bring one question your BI backlog can't answer quickly, or one cost you want to bring down. We'll show you the agent that works it out.

Book an InsightMesh demo

AI agents

The supervisor reads each question and routes it; the specialist works it out with deterministic algorithms and explains the result.

  • Supervisor

    Routes every question

    Understands the question, keeps the conversation on topic and suggests the next questions to ask.

  • Insights

    Analyzes and charts

    Answers questions about shipments and SKUs with analysis and a chart, including "why did this change?".

  • Dynamic Consolidation

    Tests shipment windows

    Finds the shipment window that saves the most, re-loading trucks against the rate card.

  • Static Consolidation

    Tests delivery days

    Tests weekly delivery-day schedules and picks the cheapest one.

  • Pallet Utilisation

    Optimizes pallets per SKU

    Tests layer patterns, pallet heights and double stacking to cut pallet count and cost.

  • Drop-Point Centralisation

    Ranks delivery points

    Ranks drop points by rate, volume or distance and compares cost and estimated CO2 before and after.

Answers you can trust

  • Deterministic numbers

    Savings and comparisons come from algorithms such as truck loading and rate-card costing; the AI plans and explains them.

  • Self-hosted model

    InsightMesh runs on a self-hosted model by default, with cloud models optional.

  • Topic guardrails

    Each assistant stays on its topic and declines questions outside it.

  • Cited document answers

    Answers from documents show the page and the passage they came from.

Integrations

Channels
Microsoft Teams Web app
Data
Excel CSV
Documents
PDF Word Excel PowerPoint Audio Images Web pages
AI models
Your private model OpenAI Claude Gemini Mistral

On the roadmap

Part of the InsightMesh architecture and in development.

  • Unified Data Access (Roadmap)

    Connect structured and unstructured data sources through a single conversational interface.

  • Knowledge Graph Intelligence (Roadmap)

    Understand relationships across customers, products, processes and business entities.

  • Role-Based Access (Roadmap)

    Keep every user to the data they are authorized to see, with row- and column-level security.

  • Enterprise Connectors & Dashboards (Roadmap)

    Live connectors to ERP, CRM, data warehouses and collaboration tools, and dashboards that refresh themselves.

Also see AIM, CortexIQ and QAI

Get started with InsightMesh

Book a demo and bring one question or one cost to bring down. We'll show you InsightMesh answering it, and the specialist agent behind the answer.

Book an InsightMesh demo

Talk to an architect

Book a demo

Let's discuss how SynthLinx can transform your enterprise.