AI implementation
We build AI that earns its budget, and connect it to the systems you already run.
Assistants, internal search, agents and workflow automation, built on your data and integrated with your CRM, website and back office. We pick the two or three uses that will pay back, choose vendors without a resale incentive, ship a working version in weeks, and measure it against a baseline.
Why it matters now
Most mid-market companies are past asking whether to use AI and stuck on what to do first. The usual result is a pile of pilots, a few new subscriptions, and no measurable change.
We have spent more than a decade on enterprise search and AI rollouts, and we build AI products ourselves. What works is simple: start with one painful, high-volume task, connect it to your real data, measure it against a baseline, then expand.
We are vendor neutral. We take no referral fees from AI vendors, and we will tell you when a tool you already pay for does the job.
What's included
Strategy and roadmap
- Use case inventory scored on value, effort and risk
- Data readiness: where your information lives and how usable it is
- Build, buy or extend-what-you-have decisions with cost estimates
- 90-day roadmap with owners and success measures
Vendor and tool selection
- Model providers and AI application vendors, compared on your use case
- AI app builders and coding tools evaluated for your team's skills
- Structured demos with your real examples, scored the same way
- Security, privacy, data retention and contract review
AI product pricing
- For software companies adding AI features
- Seat, usage and outcome-based pricing models compared
- Model and infrastructure cost per customer, modeled
- Packaging and positioning against competitors
Build and integrate
- Assistants for customer intake, FAQs and scheduling
- Internal search and question answering over your documents
- Agents and workflow automation across email, CRM and back office
- Connecting AI assistants to your systems through standard protocols such as MCP
Workflows and deployment
- AI workflow automation across intake, quoting, support and back office
- Private deployments on AWS Bedrock, Azure AI Foundry or Google Vertex AI
- Monitoring, evaluation sets and cost controls in production
- Governance and security handled alongside (see AI governance and security)
How it runs
Assess
Interviews and a review of your workflows and data to find where AI saves real hours or wins real revenue.
Choose
Pick one or two use cases, define success in numbers, and select the tools with a written recommendation.
Pilot
A working version with your data and a small group of users, measured against the baseline.
Scale or stop
Roll out what worked, document it and train the team. Stop what didn't and move the budget elsewhere.
What you get in hand
- Scored use case matrix and 90-day roadmap
- Vendor comparison with a specific recommendation
- Working pilot with measured results
- Runbook, prompts, evaluation set and training materials your team owns
- Pricing model and cost-per-customer analysis (AI products)
What we've built
Real work, described without client names.
Live AI visibility check
The free check on this site: it puts real buyer questions to Perplexity, ChatGPT and Claude with web search on, and reports whether a business was named and which sources were cited.
AI training coach
A mobile web app that builds personalized training plans with a small, fast language model, kept affordable enough to run for a growing base of real users.
Photo-to-listing agent
Identifies an item from a photo, estimates its resale value and drafts marketplace listings automatically.
Assistant connectors
Servers that let AI assistants read and write data in outside platforms, such as fitness and scheduling systems, through the Model Context Protocol.
Search quality evaluator
A test harness that scores site search results across six quality dimensions, so search and AI answer changes can be measured instead of guessed.
Pricing for this service
Start with a call or an assessment. Fixed prices for defined work, a monthly retainer for ongoing advice. Full terms are on the pricing page.
| Option | Price | What it covers |
|---|---|---|
| AI strategy call | $450 | 60 minutes with written notes and recommended next steps. |
| AI Readiness Assessment | $7,500 fixed | Two to three weeks: interviews, use case scoring, data check, vendor shortlist, 90-day roadmap, executive readout. |
| AI vendor or tool selection | $3,500 fixed | Requirements, shortlist of three, structured demos, scored comparison and written recommendation. |
| AI product pricing review | $2,500 fixed | For software companies: pricing model options, cost per customer, packaging recommendation. |
| AI pilot build | from $15,000 fixed | One use case built and measured with your data, typically four to six weeks. |
| Fractional AI advisor | $3,500 / month | About 8 hours a month of senior time, async questions, vendor and contract reviews. Month to month after 90 days. |
Common questions
Which AI vendors do you recommend?
It depends on your data, budget and risk tolerance. We work across the major model providers and application vendors, have no referral deals, and give a specific recommendation rather than a matrix.
Is our data safe?
Data handling is part of every recommendation: where it is processed, whether it is retained or used for training, and who can see it. We favor tools with business-grade data controls and will tell you when a consumer tool is a bad idea.
Should we use an AI app builder or hire developers?
App builders are excellent for prototypes, internal tools and validating an idea quickly. Production software with real customers, payments or sensitive data usually needs engineering review. We help you decide where the line is for your project.
How should we price our AI features?
Seat, usage and outcome-based models all work in the right situation. The answer depends on how predictable your model costs are and how directly the feature drives a result your customer can measure. The pricing review models it with your numbers.
Do we need a data scientist?
Rarely, for the use cases most mid-market companies start with. You need a clear owner, decent source documents and a way to measure the result.
Can you build it, or just advise?
Both. Strategy is led by us; builds are done with our long-standing development team and handed over with documentation so you are not dependent on us.
Often paired with
Not sure where AI fits yet?
Book a free 30-minute intro. Bring your list of ideas and we'll tell you which one to try first, and which to skip.