Support

Frequently asked questions

Everything about how we scope, price, and deliver AI and software projects. Can't find it here? Ask the assistant in the corner, or ask us directly.

General

How is a project priced?

Most projects are scoped as a fixed quote after a discovery call, based on what you need built. Ongoing work, such as model retraining, feature additions, and support, can move to a monthly retainer instead. Hourly work starts from $65/hr.

How long does a typical project take?

A focused chatbot or MVP web app usually takes 3–6 weeks. Full AI/data science builds or multi-tenant SaaS platforms typically run 8–16 weeks, delivered in weekly increments rather than one final release.

Do you work with our existing codebase or data warehouse?

Yes. Most engagements plug into an existing product, database, or data warehouse rather than starting from a blank repository. We audit what's there during discovery before proposing changes.

Which AI models and providers do you use?

We work across open-source models via Hugging Face and hosted models from OpenAI and Claude, choosing per use case. Cost, latency, and data sensitivity all factor into that decision rather than defaulting to one provider.

Can you build only the AI piece and leave the rest of our stack alone?

Yes. Each solution is scoped independently. You can bring us in for just the AI piece, such as a voice receptionist or a chatbot, and connect it to a system your own team already maintains.

How do you handle data privacy and security?

Every engagement starts with an NDA, and access to your systems and data is scoped per project. We can work inside your cloud environment instead of moving data into ours, if that's a requirement.

Do you offer support after launch?

Yes. Monitoring, retraining, and on-call support are available after launch, either as part of the original scope or as a separate retainer.

What size companies do you typically work with?

We work with early-stage startups building their first AI feature through to established companies modernizing legacy systems or scaling an existing SaaS product. Most clients are US small and mid-size businesses.

Do you work with US businesses across time zones?

Yes. We are remote-first and work with clients across US time zones, with a typical reply time of one business day and scheduled calls at times that suit you.

Pricing

Is the calculator result a real quote?

No. It's an indicative range based on typical rates. A real quote is confirmed after a short discovery call, once we understand the actual scope.

Can I switch between hourly and fixed-price mid-project?

Yes. Some clients start hourly to validate scope, then move to a fixed-price phase once requirements settle.

Do retainer hours roll over if unused?

Unused hours roll over for one additional month, after which they reset. This keeps monthly capacity planning predictable.

AI Chatbots & Voice Receptionists

What does an AI voice receptionist actually do?

It answers inbound calls, handles common questions using your own information, collects the caller's details, and books an appointment into your calendar. When it cannot help, it takes a message or transfers the caller to you.

Will customers know they are talking to an AI?

We design the greeting and tone with you, and we recommend being upfront that it is an AI assistant. Callers who need a person can be handed off, so the experience stays helpful instead of frustrating.

Which systems does it connect to?

Typically your calendar, your CRM, and email or SMS. Common choices are Google Calendar, HubSpot, and Twilio for phone numbers, and we can connect other tools through their APIs or automation platforms such as n8n.

How long does a chatbot or voice agent take to build?

A focused website chatbot usually takes three to six weeks, delivered in weekly increments. Builds with several channels and integrations take longer, and we confirm the timeline after the discovery call.

Explore AI Chatbots & Voice Receptionists

Agentic AI & AI Agents

What is agentic AI?

Agentic AI is software that can take a goal, plan the steps, use tools such as your CRM, email, or databases, and carry out the work, rather than only answering a question. It works within limits you set and asks a person when it should.

How is an AI agent different from a chatbot?

A chatbot answers questions. An agent completes tasks: it can look things up, update records, draft and send messages, and move work forward across several systems.

Is it safe to let an agent act on our systems?

We give each agent only the permissions it needs, add approval steps for anything risky, and log every action. You decide what it can do alone and what needs sign-off.

Which frameworks and models do you use?

Typically LangGraph or CrewAI with models from OpenAI, Anthropic Claude, or open-source models, chosen per task for accuracy, cost, and data privacy.

Explore Agentic AI & AI Agents

RAG & Knowledge Assistants

What is RAG?

Retrieval-augmented generation is a way of building AI assistants that first search your own documents and data for relevant information, then use a language model to write an answer based on what was found, with citations.

Will the AI make things up?

RAG greatly reduces this because answers are grounded in your content. We add citations, accuracy testing, and guardrails so the assistant says it doesn't know when the answer isn't in your data.

Can it respect who is allowed to see what?

Yes. Retrieval can be filtered by the user's permissions, so people only get answers from documents they already have access to.

Can RAG run on Databricks or Snowflake?

Yes. We can build retrieval on Databricks Vector Search or Snowflake Cortex Search so your data stays inside the platform you already use.

Explore RAG & Knowledge Assistants

Business Process Automation

Which processes are worth automating first?

The ones that are frequent, rule-based, and painful: copying data between tools, sending follow-ups, building weekly reports, and processing documents. We map these in the discovery call and start with the highest time saving.

Do you use no-code tools or custom code?

Both. n8n, Make, or Zapier are fast and easy to maintain for many workflows. For complex logic, high volume, or systems without APIs, we write custom code.

Can automation work with our existing software?

Usually yes. Most modern tools have APIs, and for older systems we can use browser automation or scheduled exports.

What happens when something goes wrong?

Every workflow has error handling, retries, and alerts, and unusual cases are routed to a person instead of failing silently.

Explore Business Process Automation

Databricks & Snowflake Data Engineering

Should we choose Databricks or Snowflake?

Both are excellent. Databricks is strongest for data engineering, machine learning, and large or unstructured data. Snowflake is strongest as a simple, SQL-first warehouse for analytics. We recommend based on your team, data, and budget, and we work with both.

Can you migrate us from spreadsheets or an old database?

Yes. We move data from spreadsheets, MySQL, Postgres, SQL Server, or older warehouses into Databricks or Snowflake, with checks that numbers match before you switch.

How do you keep platform costs under control?

Right-sized compute, auto-suspend, incremental pipelines, and cost monitoring dashboards, so you pay for what you use.

Can our data be used for AI and RAG?

Yes. We prepare governed, AI-ready tables and set up vector search on Databricks or Snowflake, so assistants and agents can use your data securely.

Explore Databricks & Snowflake Data Engineering

CRM & Follow-up Automation

How do automatic follow-ups work?

You set the rules once, for example sending a reminder if a lead has not replied in three days. The system sends the message from your own email account and stops the sequence as soon as the lead responds.

Can it replace the CRM we already use?

Sometimes. If your current tool works, we can add follow-up automation on top of it. If it does not fit how you sell, we build a CRM around your actual process and migrate your data.

Do the emails come from our own address?

Yes. The system connects to your email provider, such as Gmail or Microsoft 365, so messages come from your normal address and replies land in your normal inbox.

Can it find and research new prospects?

We can add AI agents that research target companies and draft personalized outreach for you to review. You stay in control of what is sent.

Explore CRM & Follow-up Automation

Data Dashboards & AI Insights

Can you build a dashboard from our Excel or CSV files?

Yes. We can start from the files you already use and set up a dashboard that refreshes when new data arrives. Later it can connect directly to a database or business tool.

What is an ask-your-data chat?

It lets your team type a question in plain English, such as which service earned the most last quarter, and get an answer with the chart behind it.

How accurate are the forecasts?

Accuracy depends on how much clean history you have. We measure forecast error against past data and show it on the dashboard, so you know how much to trust each prediction.

Is our data safe?

Every engagement starts with an NDA, and access is limited to what the project needs. If required, we can work inside your own cloud environment instead of moving data elsewhere.

Explore Data Dashboards & AI Insights

AI-Powered Websites & Web Apps

Can you add AI to our existing WordPress site?

Yes. We build custom plugins and add features such as chat assistants and instant quote forms without rebuilding the whole site.

What is an instant quote tool?

Visitors answer a few questions and see a price range straight away, and you receive their details with the answers attached. It filters serious buyers and saves quoting time.

How long does a website or web app take?

A focused marketing site or MVP web app typically takes three to six weeks. Larger apps with accounts and integrations take longer and are delivered in weekly steps.

Will the site be fast and findable on Google?

We build for speed and clean structure, with proper titles, descriptions, structured data, and mobile layouts, which gives your pages a solid base for search.

Explore AI-Powered Websites & Web Apps

SaaS Products, Ready to Sell

What does multi-tenant mean?

One application serves many customers while keeping each customer's data separate. It is the standard design for SaaS and is much harder to add later, so we build it in from the start.

Can you handle subscriptions and billing?

Yes. We connect plans, free trials, upgrades, and invoices to a payment provider such as Stripe and tie them to what each customer can use.

How long does a SaaS MVP take to build?

Multi-tenant SaaS platforms typically run eight to sixteen weeks, delivered in weekly increments. We agree a fixed scope after the discovery call.

Who owns the code?

Ownership of the code is set out in your project agreement, and we build on open, current tools your own team can maintain, so you are not locked in.

Explore SaaS Products, Ready to Sell

Mobile Apps with AI Built In

Do you build for both iPhone and Android?

Yes. Most projects use a single cross-platform codebase that ships to both stores, and we use native development when performance or platform features require it.

Can the app work without internet?

Yes. We can design offline-first apps that store work on the phone and sync when the connection returns, which is useful for field teams.

What AI features can an app include?

Assistants, photo recognition, and smart suggestions are common. Depending on cost and privacy needs, they can run on the device or call an AI service through an API.

How long does a mobile app take?

A focused first version is scoped in weeks rather than months, and we agree a timeline after the discovery call, with builds you can test on your own phone every week.

Explore Mobile Apps with AI Built In

Related solutions

More ways we can help

Hand-picked solutions that work well together. Hover a panel to see what's inside, or browse all ten.

All 10 solutions
02

Solution 02

Agentic AI & AI Agents

AI agents that plan a task, use your tools and data, and finish multi-step work on their own, with guardrails and a human in the loop where it matters.

  • Single-task AI agents
  • Multi-agent workflows
  • Human-in-the-loop guardrails
  • Monitoring & evaluation
Explore solution
  • Multi-step tasks done
  • Uses your tools
  • Human approval

Still have a question?

Send it over. We typically reply within one business day.