01 ADVISE
AI strategy & opportunity review
We look at your processes, data, systems and goals to find where AI can create practical value, and where a simpler fix is the better answer.
AI Opportunity Review
$1,500 · fixed fee
2 weeks
Blueberry Intelligence & Automation
From AI opportunity to operational impact.
We help SMEs and NGOs identify where AI can create real value, build practical systems, improve existing ones, and equip their teams to use them effectively.
Or email ada@blueberryia.com. We reply within two business days.
Quick question? . It is the same Website Support Assistant we install for clients.
The problem
AI is easy to discuss. Making it useful is harder.
Organizations do not need more AI hype. They need clear decisions about where AI belongs, systems that work reliably, and teams that know how to use them well.
Unclear where AI belongs
Plenty of ideas, no clear way to tell which are worth funding and which are better solved with a simpler fix.
Demos that do not hold up
Pilots that impress in a meeting but give wrong answers, cost too much or break once real users arrive.
Tools without new habits
Staff have AI tools, but the way work is done has not changed, so the benefit stays small and uneven.
Advise → Build → Optimize → Enable
Four ways we help, with clear starting prices.
Start with the part where you need help most. Each service has a fixed-scope first engagement, so you know the cost before you commit.
02 BUILD
AI & intelligent systems
We design and build AI applications, knowledge assistants, automations, integrations and the supporting software around them.
Pilot Build
From $6,500 · quoted after scoping
4–6 weeks
03 OPTIMIZE
AI reliability & improvement
We improve retrieval, grounding, evaluation, performance, cost and maintainability in AI systems that already exist.
AI System Health Check
$1,800 · fixed fee
1–2 weeks
04 ENABLE
AI adoption & team capability
We train teams, redesign workflows and help people adopt AI responsibly in their day-to-day work.
Team AI Workshop
$1,200 · per half-day · $3,800 for 4 sessions
Half-day, or a 4-session programme
Most engagements are paid 50% upfront and 50% on delivery. See full scope and terms.
Build · ready-to-deploy
Website Support Assistant
Visitors get quick answers about your services, prices, hours and policies, drawn only from documents you approve. When it does not know, it says so and offers one-tap WhatsApp, email or call buttons to reach your team.
You can try it right here. The assistant on this website is the same product we install for clients, set up with Blueberry's own documents.
For SMEs
$1,200
one-off build
For NGOs
$1,000
one-off build
- Terms: 50% upfront, 50% at go-live.
- Running costs: You own it. Hosting (a small cloud server) and AI usage are billed to you directly by the providers, with no per-message platform fees and no lock-in.
- Support: Optional care plan: $100/month for updates and fixes. Not required.
What it does
- Answers only from the documents you provide (PDF, Word, Excel/CSV, PowerPoint, HTML, Markdown or text) and shows the sources behind each answer
- Finds exact details reliably, such as prices, product codes and phone numbers, using keyword plus semantic search
- Hands over to your team with WhatsApp, email and call buttons when the answer is not in your documents
- Added to your website with a single script tag, styled with your colors, logo, greeting and starter questions
- A password-protected admin page to upload or remove documents and test answers
- One private instance per client: your documents and chat history stay on your own server
It answers questions and hands over to people. It does not take bookings or payments, and it is not a WhatsApp bot.
How we work
Start with the problem, not the technology.
Every engagement should leave you with more than a prototype: a useful system, a clear way of running it, and the capability to sustain it.
-
Understand
Clarify the real problem.
We look at the workflow, the data and the people involved before choosing any technology.
-
Design
Choose the simplest approach that works.
Clear system boundaries, a build-vs-buy decision and agreed ways to check quality.
-
Build
Make it work on real data.
Applications, knowledge pipelines and integrations with tests and documentation.
-
Improve
Measure, then refine.
Check answer quality, speed and cost against real questions, and fix what fails.
-
Transfer
Leave your team able to run it.
Runbooks, training and handover so the system does not depend on us.
Who we serve
Built for organizations doing real work.
SMEs
Practical technology and AI without enterprise-scale overhead. Typical starting points: answering customer questions faster, cutting repetitive admin, and making internal knowledge easy to find.
NGOs & development organizations
Responsible, maintainable systems designed around operational realities, limited budgets and the people who will run them, from programme data and reporting to staff knowledge and supporter enquiries.
Selected systems
Proof over promises.
Systems built by our founder, described as problem, approach and system. Client names and results are only published with permission.
Public health DHIS2 integration In development
Public health data assistant (DHIS2)
- Problem
- Public health teams spend a lot of time running the same indicator queries in DHIS2 and turning the results into reports.
- Approach
- A conversational layer that turns plain-language questions into structured requests to the DHIS2 Analytics API, using validated indicator definitions, with a deterministic fallback when the AI is unsure.
- System
- An analytics assistant that answers indicator questions from DHIS2 data and produces tables, charts and report-ready summaries.
- Python
- FastAPI
- DHIS2 Analytics API
Internal operations Role-based access Built
Secure internal knowledge assistant
- Problem
- Operational knowledge was scattered across drives and manuals, so senior staff kept answering the same policy questions.
- Approach
- A retrieval-augmented assistant with role-based access control, so each person only gets answers from documents they are allowed to see, plus routing of unresolved questions to the right team.
- System
- An internal assistant that answers from authorized documents with citations and turns questions it cannot answer into routed support tickets.
- Python
- FastAPI
- LangChain
- Role-based access control
Customer support Hybrid search Built
Multi-tenant customer support platform
- Problem
- General-purpose chatbots gave confident but wrong answers and could not keep each organization’s rules and documents separate.
- Approach
- Strict tenant isolation, hybrid retrieval (keyword + dense vector search) and local re-ranking, so answers come only from the right organization’s documentation.
- System
- A support platform that turns each organization’s documentation into grounded answers and hands complex requests to people.
- Python
- FastAPI
- Hybrid search
- Re-ranking
- Next.js
- TypeScript
Governance & operations Local transcription Built
Privacy-first meeting minutes
- Problem
- Organizations wanted structured minutes from confidential meetings without sending raw recordings to a cloud service.
- Approach
- Transcribe audio locally, remove personal information on the device, and only then use a language model to structure the redacted text.
- System
- A meeting tool that produces minutes, decisions and action owners while keeping raw audio and unredacted transcripts on the machine.
- Python
- Whisper (local)
- spaCy
- Gemini API
Insights
Practical thinking about AI.
Clear perspectives on AI strategy, implementation and adoption, written for people who make decisions.
Strategy & Decision-Making
What should an SME actually automate first?
In a 2025 OECD survey of more than 5,000 SMEs across seven countries, only 29% of AI users applied it to core activities. Start not with the most exciting tool, but with the most boring task that passes a four-question filter.
· 5 min read
Engineering & Implementation
What makes an AI system reliable?
When Air Canada's chatbot misstated bereavement fare rules, a Canadian tribunal ordered the airline to pay. What separates an AI assistant you can trust from one that quietly makes things up?
· 6 min read
Adoption & Enablement
Why AI training without workflow redesign often fails
In McKinsey and BCG studies, the single factor with the largest impact on AI financial returns was process redesign. Training people on a tool while leaving their workflows unchanged usually leads to the tool being dropped.
· 5 min read
Questions
Before you get in touch
Short, honest answers to what organizations usually ask first.
Do we actually need AI?
Not always. Some problems are better solved with a simpler workflow, a standard tool or ordinary software. Part of our job is to tell you when that is the case, before you spend money on a build.
Who do you work with?
Small and mid-sized businesses (SMEs) and NGOs and development organizations. We work remotely with clients in different countries, and all of our work is delivered in English.
What does it cost?
Our starting engagements have fixed prices: an AI Opportunity Review is $1,500, an AI System Health Check is $1,800 and a half-day team workshop is $1,200. Pilot Builds start from $6,500 and are quoted after scoping. The Website Support Assistant is a one-off $1,200 for SMEs or $1,000 for NGOs.
How do payments work?
Most engagements are 50% upfront and 50% on delivery. Pilot Builds are paid 40% upfront, 30% at an agreed midpoint and 30% on delivery. The Opportunity Review fee is credited in full against a Pilot Build signed within 60 days.
How do you handle our data?
We only ask for the information a piece of work needs, use it only for that work, and are happy to sign an NDA. Where it makes sense, systems run on infrastructure you control. Please do not send passwords or sensitive personal data through the contact form.
Is the chat assistant on this site real?
Yes. It is our own Website Support Assistant, the same product we install for clients, answering from Blueberry’s own documents. If it cannot answer, it points you to our email.
Who will we work with?
Blueberry was founded by Victor Kola. You work directly with the people designing and building your system, not through layers of account management.
What happens after delivery?
You own what we build. Every build ends with documentation and a handover session, so your team can run it. Ongoing support is available if you want it, but it is never compulsory.
Have an AI problem worth solving?
Tell us what you are trying to improve. We will help you work out whether AI, automation, software or a simpler approach is the right answer.