Question 1

Are you Sydney-only?

Accepted Answer

We are based in Sydney and work with organisations across Australia. Delivery is a mix of on-site workshops and remote execution, depending on your needs.

Question 2

Which systems do you integrate with?

Accepted Answer

Common anchors include HubSpot, Zoho, Salesforce-style CRMs, Xero/MYOB class finance tools, Google Workspace / Microsoft 365, ticketing and chat tools, WhatsApp/SMS/email providers, Monday/Asana/Notion, and bespoke line-of-business databases. We pick integration patterns based on your existing licences, APIs, and compliance profile.

Question 3

Can you work alongside our internal team?

Accepted Answer

Yes. We routinely work with owners, operations, finance, and IT — shared checkpoints, documented assumptions, and handover packs so you are not dependent on a single vendor brain.

Question 4

How is this different from a cheap "AI chatbot"?

Accepted Answer

Low-cost chatbots are often thin wrappers with little integration, weak guardrails, and no accountability for your P&L. Our engagement programmes target measured pilots: named workflows, CRM or finance touchpoints, logging, human review where risk demands it, and acceptance tests you sign off. Your proposal spells out risk transfer, outcomes, and commercial terms after discovery.

Question 5

How do you handle accuracy, privacy, and compliance?

Accepted Answer

We default to human-in-the-loop patterns for high-risk decisions, role-based access, audit trails, and data minimisation. Regulated sectors (health, finance, government) include explicit governance items in scope from the start.

Question 6

Do you still build websites or only AI?

Accepted Answer

This microsite emphasises operational Data & AI. Yarli Data still delivers web products where they unlock the workflow (forms, portals, customer-facing surfaces). If your primary need is a marketing site, say so in enquiry and we will route you to the right programme.

Question 7

What is written acceptance for an AI pilot slice?

Accepted Answer

Written acceptance is a short checklist sponsors sign each cycle: inputs, outputs, error handling, notifications, and human override. It keeps scope visible in review cadence. See https://yarli.com.au/blog/written-acceptance-ai-pilot-slices

Question 8

What is a grounded internal knowledge pilot?

Accepted Answer

It is a bounded internal Q&A programme using approved sources, citations, role-based access, and clear escalation. Start with one operational slice. Guide: https://yarli.com.au/blog/grounding-internal-knowledge-pilot

Question 9

Why do pilot demos need a staging environment?

Accepted Answer

Staging should mirror production permissions and data shapes so sponsors judge real risk. Laptop-only demos hide access and PII gaps. See https://yarli.com.au/blog/staging-environment-pilot-reviews

Question 10

What is human-in-the-loop for customer-facing AI?

Accepted Answer

A person approves, edits, or rejects automated drafts before customers see them, with logs and override playbooks. Details: https://yarli.com.au/blog/human-in-the-loop-customer-facing-ai

Question 11

When should we clean CRM data before automation?

Accepted Answer

Before connectors and workflow automation scale, define one-row rules, merge logic, and field ownership. See https://yarli.com.au/blog/data-quality-before-models and https://yarli.com.au/blog/crm-deduplication-before-automation

Question 12

What is a review cadence?

Accepted Answer

Review cadence is the fixed rhythm (usually weekly or fortnightly) at which a pilot team demos real progress, records exceptions, and updates written acceptance so sponsors see runnable work. Full guide: https://yarli.com.au/blog/review-cadence-data-ai-pilots

Question 13

What is data cadence?

Accepted Answer

Data cadence is how often operational feeds refresh, quality checks run, and golden scenarios replay so automation and retrieval stay trustworthy. It pairs with delivery review cadence in a pilot. See https://yarli.com.au/blog/what-is-data-cadence

Question 14

How often should we review a Data or AI pilot?

Accepted Answer

Most teams use a weekly or fortnightly 45-minute session with an operational sponsor, technical lead, and someone who lives the workflow. Fortnightly is fine when dependencies are stable; the slot should stay fixed rather than when ready. Details: https://yarli.com.au/blog/review-cadence-data-ai-pilots

Question 15

Do you offer Python bootcamps?

Accepted Answer

Our public programmes focus on operational Data and AI enablement for CRM, documents, workflows, and governance. We do not market introductory Python bootcamps on this site. See https://yarli.com.au/trainings

Question 16

How are projects scoped and priced?

Accepted Answer

We start with a short discovery, then propose either a fixed-scope pilot (clear acceptance tests) or a time-and-materials slice with a not-to-exceed guardrail. Retainers are quoted once we know monitoring cadence and stack complexity.

Question 17

Do you train our team on the new workflows?

Accepted Answer

Yes. Enablement is part of adoption: playbooks for edge cases, how to override the automation safely, and short sessions for frontline staff — not just a technical handover to IT.

Question 18

How do we get started?

Accepted Answer

Email info@yarli.com.au or use the contact form. A short intro call helps us understand priorities, constraints, and timelines before we suggest next steps.

Question 19

How long does a typical pilot take?

Accepted Answer

Depends on integrations and data readiness. Many Discovery & Pilot Programme engagements run 4–8 weeks once access and stakeholders are aligned; Revenue & Ops Intelligence Build commonly 8–12 weeks; Workflow & Orchestration Programme and Tier 4 work runs longer with phased milestones. Dates are fixed in your proposal at kick-off.

Question 20

We already have Zapier / make.com — can you help?

Accepted Answer

Yes. Low-code tools are a great start until logic sprawls, errors are hard to trace, or you need governed AI on top of the data they move. We audit what exists, harden critical paths, and propose either incremental fixes or a structured replatform — commercial minimum depends on what we find.

Question 21

How do the Discovery & Pilot, Revenue & ops intelligence, and Workflow programme tiers differ?

Accepted Answer

Discovery & Pilot Programme: one high-friction workflow (often internal knowledge or document handling), mapped state, 1-2 integrations, playbooks, and hypercare as scoped. Fixed proposal after discovery. Revenue & Ops Intelligence Build: CRM intelligence plus quote/ticket lanes and reporting, with a second workflow hook-up as scoped. Workflow & Orchestration Programme: multi-system programmes with KPI-led dashboards and change management. ERP & custom decision-support AI: deeper ERP or multi-entity orchestration, scoped and quoted after discovery. Programme detail on the Work page.

FAQ

Frequently asked questions