AI features in your product
LLMs in real applications: content generation, classification, data extraction, chat over documents. From prompt to production API - in your stack.
AI consulting · AI / LLM implementations for companies
Anyone can spin up an AI demo. Almost no one ships it to production.
I help companies get from "we want to ship AI" to a solution running in production - with costs you can budget, results you can measure, and a process your team can run without me. No hype, just specifics.
Scope
What I actually help with
Not a "digital transformation strategy". Working code, measured results, and a team that knows what it's doing.
RAG and chat over company knowledge
Retrieval, embeddings, chunking, reranking - explained in plain language and implemented so answers come from your data, not the model's imagination.
Agents and automation
Agentic workflows in products and in the dev process: subagents, MCP, hooks, orchestration. Where automation actually pays off - and only there.
Evals and quality control
How do you know the AI works? Test sets, LLM-as-judge, quality gates in CI. Metrics instead of "well, looks fine to me".
Token costs under control
Model selection, caching, request architecture, spend monitoring. So the API invoice stops being a monthly surprise.
AI in the dev process (AI-SDLC)
Claude Code and agents in your team's daily work: standards, code review, security, measuring impact. A process, not an experiment.
Process
How the engagement works
The same pipeline you saw above - just in more detail. Every phase ends with something you can actually touch.
- 01
Audit
1-2 weeksI review your ideas, projects, and data. I map use cases from "quick win" to "ambitious but risky". I estimate token costs and risks before anyone writes a line of code.
What you get: a report with a use-case map, cost estimate, and pilot plan
- 02
Pilot
2-6 weeksWe take the smallest sensible use case and build a working prototype in your stack. Evals from day one - we know whether it works from the start, not whether it "looks nice in a demo".
What you get: working code in your repo + quality and cost metrics
- 03
Production
depends on scopeHardening the pilot: monitoring, guardrails, error handling, data security, cost control. All the boring parts that separate a demo from a product.
What you get: a solution in production with monitoring and a token budget
- 04
Handoff
1-2 weeksDocumentation, pairing with your team, training on maintaining and extending the solution. My goal: a team that doesn't need a consultant on retainer.
What you get: a team that ships on its own + documentation in the repo
Fit
Does this make sense for you?
I'd rather say "no" on a call than take a project that can't succeed.
Good time to call, if:
- you know what AI should improve in your business - what's missing is someone who knows how to ship it
- the demo exists, but you're afraid to push it to production
- API costs are growing faster than the value you can point to
- your team knows its craft inside out, but LLMs are new terrain
- you want to build the capability in-house, not rent a consultant forever
Bad time to call, if:
- you're looking for someone to "do the AI" without your team involved
- there's no use case - just a "we need AI" memo from the board
- you expect a guarantee the model will never be wrong
- the project mostly needs classic ML research, not an LLM implementation
Not sure which side you're on?
30 minutes and you'll know. If I can't help - I'll say so straight and point you to someone who can.
Calculator: what does waiting on your AI rollout cost?
Every month without a working solution is value left unrealized. See what postponing the rollout costs you - and what you recover by starting sooner.
Total cost of delaying the rollout:
vs the cost of consulting - I'll show you the actual numbers on the call.
Book a 30-min call →A ballpark estimate - the real numbers depend on your use case and organization.
Tomasz Guściora
Claude Code Trainer & Hands-on Practitioner
13 years in the industry - most of it in data science. 750+ hours in Claude Code, since back when something broke every week (and I mean every week). Production AI/ML I've shipped: fraud detection, customer segmentation, proposal generators, startup prototypes. I'm not a trainer-of-trainers. I teach what I actually do at the keyboard every day.
I understand LLM architecture not from articles - from banging my head against it during real implementations. That's why I can explain to your seniors why Claude Code occasionally "hallucinates" and, more importantly, how to write code that keeps that risk on a leash.
Substack DemystifAI - three years of poking at large language models: token economics, agents, AI in the code-generation loop. GitHub claude_code_common_base - my own Claude Code template, with hooks, subagents, MCPs. Audit the code before you book the call. Seriously.
6 years at Citi (Warsaw, London, Jakarta) taught me how big orgs actually work: compliance, governance, procurement, NDA, GDPR. I know what derails an adoption in a software house - and it's never "just the technology". It's people. It needs a shift in how the team thinks. I'll help you with both.
See my GitHub →Common questions about consulting
How is consulting different from the training?
Training builds your team's skills - I teach you how to work with Claude Code and agents. Consulting is shipping a specific solution: we sit down with your project and deliver working code to production.
In practice the two often combine. First an implementation pilot, then team training so the solution keeps living without me. The order can flip too - depends on what hurts more.
Do you write code or just advise?
Both - with a strong lean toward code. The audit ends with a report, but the pilot and production phases happen in your repo: commits, pull requests, code review. Usually pairing with someone from your team, because the knowledge should stay with you, not with me.
I'm not a slide-deck consultant. What's left after the engagement is working code and people who know how it works.
Our stack is .NET / Go / Rust. Is that a problem?
No. The AI layer - model APIs, RAG, evals, agents, token costs - is language-agnostic. I personally write mostly Python and JavaScript, but the principles work the same everywhere. During the pilot I work in your stack, pairing with your developer - which happens to be the fastest way to keep the knowledge in your team.
What about NDAs, GDPR, and data security?
Standard practice. I sign an NDA before looking at any code. Six years at Citi (Warsaw, London, Jakarta) taught me how compliance works in large organizations - GDPR, governance, and procurement are routine for me, not obstacles.
On the architecture level: we match the solution to your requirements - from APIs with DPA agreements to models hosted in your own infrastructure. It's one of the first things we settle in the audit.
What does it cost and what formats do you work in?
Three formats: audit (fixed price), pilot/implementation (quoted after the audit), and ongoing advisory (monthly retainer). The exact quote depends on scope - which is why we start with a 30-minute call, no strings attached. If you like what you hear - I'll prepare and send an offer.
Book
Grab a free 30 min
A short discovery call - we figure out if the training fits, and you get a few useful pointers either way.
30 min
- We dig into what your team actually needs
- Every question gets a straight answer
- No pitch, no pressure, no obligation
Zero risk - it's just a conversation
Open in new tab Rather type than talk? → Drop me a message
Looking for something other than consulting?