I’ll find the routine that’s costing you money — and hand it to an agent
I’m Vlad Derkach, Gdańsk. I run the review, I build the agent, and I maintain what I built. who I am →
First I go through your processes and show in numbers where money is leaking — that part is free. Then I take one routine and lift one number in it. Upfront payment is zero. You pay only a share of the uplift — and only once the number has grown and a second measurement shows it.
Free review: where routine work is costing you A review is a map of your repeating work, with the price of each one in hours and money. Message me on Telegram — I answer within the day. The review commits you to nothing: the numbers are yours to keep. Not ready to talk — go through the routine map yourself: 18 items, ten minutes, no form and no email.Works with what you already use
The agent sits in your messenger, your inbox and your spreadsheets. There’s nothing new to learn — not for you, not for your office.
Need something else — the inbox where enquiries land, your website, a stock spreadsheet, a payment export to reconcile? Ask. I’ll tell you honestly whether it connects or not — before I take the job on, not after.
Routine work often lives in messages — and gets lost there too
Enquiries arrive in your messenger, your inbox and your spreadsheets. Evenings and weekends go unanswered, reminders depend on somebody remembering, price checks get put off. The team spends more time managing the work than doing it.
What has already been counted
I label every number with its honest status — the same way I do in client reports. You won’t find promises of growth “several times over” here.
That’s how many checkable facts (prices, links, promotions) my agent verifies every week on a paid travel-guide site. It has surfaced 26 discrepancies, each with evidence; six were critical, including a discount promised in two different cities at once.
✅ fact — running in productionAn estimate of what a landlord with six flats loses to a manual monthly cycle: hours of routine, errors in utility reconciliation, late payments. Three scenarios; the rates and some of the figures are market assumptions.
🟡 hypothesis — awaiting measurementA sector-wide measurement will go here: I’ll go through 10 companies as a mystery shopper and publish how many of them lose enquiries sent in the evening. Your company can be one of the ten — free of charge.
📌 plannedThree steps. I’m the one risking money, not you
We start with one routine. We prove the number. We expand — if there’s anything to expand.
Free review
I go through your processes and find the routine that costs the most: how many hours it eats, what gets lost, what that is in money. Where it helps, I check as a mystery shopper — in the evening and at the weekend. We put the starting point in writing — so there’s nothing to argue about later.
It costs zero. The numbers are yours even if we don’t work together after that.
A week of work
I put an agent on the chosen routine: it does the work itself along rules we’ve agreed — replying, cross-checking, counting, reminding. Anything the agent isn’t sure about goes to a human. The week looks like this: the review and the measurement, two days of building, a day to agree the scenarios and the limits, then launch on live enquiries.
Your time on “implementation” — just answering questions as they come up.
Paid from the uplift
After the agreed period we run the review again, by the same method. If the number reached the threshold, you pay the agreed share of the uplift — for three months from the agent’s launch. After that I hand the system over to you and the payments stop.
If it didn’t grow, you pay nothing.
Your time, counted honestly
What repeating work costs you now — and how the arithmetic changes when an agent takes it over. Upfront in this model is always zero: I earn only on the uplift, so I count it in the open.
I take one routine at a time. The agent takes three quarters of its hours off you; a quarter stays with a human — for checking, and for everything that falls outside the script.
- Taken over by the agent on this routine
- Left to a human
- Saved per month on this routine
- Upfront$0
- Your uplift over 3 months
- My share (% for three months from launch)
- Your net gain over 3 months
Take all 3 routines later — potential a month. 🟡 hypothesis
An indicative calculation from the numbers you entered, not a promise. The maths: hours × people × cost of an hour × 4.3 weeks — that is what all your repeating work costs now; then divide by the number of routines, the agent takes three quarters of one routine’s hours, the uplift over three months, minus my share. The starting point, the threshold, the share and the term go into writing at the free review — on your real numbers, not these.
One operator for repeating work
The routines that automate well are the ones that repeat daily and live in messages, calls or spreadsheets.
Incoming enquiries
Works out the level and the goal, suggests a course, a service and a date, moves the person to a booking. Answers in the evening and at the weekend.
Reminders and bookings
Keeps the schedule, reminds the customer and the office, picks up a stalled enquiry before it goes cold.
Content and price control
Once a week it cross-checks prices, links and promotions on the site. It sends only the discrepancies — each with evidence.
Reconciling data and payments
Matches spreadsheets, hunts for discrepancies, tracks late payments. It never touches money — it only shows you where the figures don’t add up.
Market monitoring
The same method I used to run a weekly review of 17 competitors for a marketing-analytics platform.
The “before → after” report
A second measurement by the same method as the starting point. My fee — and your gain — are counted from that number.
Not screenshots — working systems
Every agent can be seen in action — not in screenshots, but at a live address. Ask anything, try to throw them off.
The most important scenario is the failure scenario
Every agent eventually meets a question that isn’t in its script. The difference between a toy and a working system is what happens next.
▶ short video: the agent gets an off-script question and hands it to a human — coming soon- The agent doesn’t touch money. No payments, no price changes, no confirming refunds — not without a human approving it.
- Anything unclear goes to a human. A question outside the agreed scenarios isn’t guessed at; it is passed to you or to your office, flagged.
- Every action is logged. What the agent said, to whom and when — you can open the log at any moment.
In the open: zero now, a share of the uplift later
A week to launch · the starting point in writing · the share and the term agreed before work starts.
$0
Free review
A map of your repeating work, with the price of each one in hours and money. The numbers are yours even if we don’t work together after that.
$0 upfront
First agent
A fixed sum — payable only once a threshold agreed in writing is reached (for example: “occupancy is up by 10%”). If it doesn’t grow, you pay nothing. Instead of part of the fee I take a written reference and permission to publish the case with its numbers.
20–30%
From the uplift
A share of what’s gained or saved — three months from the agent’s launch, and that’s it. After that I hand the system over to you and the payments stop: the percentage never runs forever. I maintain what I built, if you want that after the handover — usually $200–300 a month.
from €4,000
AI product audit
A different practice, for product teams: where AI actually pays inside the product and operations, which agents to build first, and what each one costs per user per month. Two weeks. After that — a fleet of agents from €12,000, or a fractional AI Product Lead from €5,000 / month.
Three things — and none of them are about money
- Access to customer conversations — connecting the bot to your business messenger (Telegram / WhatsApp) or to the inbox where enquiries land.
- 30–40 minutes talking through the process: how enquiries arrive today, who answers them, what counts as success.
- Your price list and schedule — what you already send customers becomes the agent’s knowledge base.
- Access to your bank, till or payment systems.
- Passwords to your CRM or admin panels at the start — if they’re needed later, we agree that separately and in writing.
- Your time for “implementation”: for a week the work is mine; you answer questions as they come up.
What the agent cannot do — and how to switch me off
- Spend or transfer money.
- Change prices, confirm refunds or enter into contracts.
- Answer outside the agreed scenarios: an off-script question goes to a human rather than into improvisation.
- Delete your conversations or your data.
- Everything built is yours: the code, the bot, the knowledge base and the logs stay with you even if we part ways.
- Access is revoked in a minute: one access key — you switch it off yourself and the agent stops immediately.
- The log of everything the agent does is open to you at all times.
- The agent and its logs run in an isolated environment in the EU. If we part ways, I hand them over to you or delete them — your choice.
- Business data is never sold or published. Only the necessary minimum reaches the language model — on API terms, with no model training on your data.
- I’ll say “no” if the task isn’t one for AI — before I take it on.
What people who have worked with me say about me
These are references for product work, not reviews from clients of the agent practice: the agent practice has no client cases with numbers yet — the first one will appear right here, in “numbers”. The names are real, and all three are verifiable on LinkedIn.
“The way he frames problems now — that is strategic-level thinking.”
Vlad Zhovtenko · CEO, RedTrack.io“He’s one of those people who turn complex, risky startup hypotheses into working businesses.”
Dzmitry Khudy · CEO, Zorka.Agency“When budgets were cut at Roketo, Vlad brought in grants, kept the team together and ran our pivot as CPO. An extremely dependable product leader.”
Taras Dovhal · Co-Founder, NoVPS; previously RoketoProduct manager who builds the agents too
Vlad Derkach — product manager and builder of AI agents. Analysing the business in numbers, researching the processes and building the agent itself — that is one person, not three subcontractors. Whoever answers your message on Telegram runs the review, builds the agent and answers for the result.
- 25+ agents in production use at a marketing-analytics platform (2025–2026) — plus 11 automated routines there, from weekly monitoring of 17 competitors to integration-error control.
- Three apps of my own in the App Store and Google Play — from the first line of code to release, alone.
- Product Discovery Flow, an open tool — a pipeline of 15 agents for product research; source code open under the MIT licence, so the method can be checked line by line.
- Before product, a lawyer: a master’s in law from BSU and two years in a bank, as legal counsel and then senior legal counsel. That is where the order I still keep comes from: evidence first, claim second.
- No employees and no subcontractors: no markup for a “team”, no broken telephone, and no one else’s projects queued ahead of yours.
- I work from Gdańsk and take one routine at a time — which is why I sometimes say plainly, “not right now”.
Check me instead of taking my word for it
There isn’t a single screenshot below, and not one link back to my own words. Only addresses that open right now — apps in the stores, source code and a working bot. The links lead outwards, to platforms I don’t control.
Three apps released in 2026. Each one passed Apple’s and Google’s review — an outside check, not my word about myself. All three are free and open without registration.
- Tapset A strength-training log: one tap per set. App StoreGoogle Playtapset.fit
- Shashlyk Assistant A helper at the grill: how much charcoal, how long, how many portions. App StoreGoogle Playmangal-time.shop
- HandoverBoat Yacht handover in charter: recording the state of the boat in verifiable form. App StoreGoogle Playhandoverboat.com
Product Discovery Flow — a pipeline of 15 agents that carries product research from a raw brief to a finished report. It installs with one command. The code is open under the MIT licence: you can see exactly what each agent does, rather than “trust the black box”.
- Open the tool The write-up and one-command install: what all 15 agents do, step by step. product-discovery.dervlad.com
- Source code The whole repository, MIT licence — read it, copy it, use it yourself. GitHub
- In a public catalogue The same tool in a third-party catalogue — proof it doesn’t only live on my own disk. skills.sh
The demo bot runs on made-up data for a sailing school: ask it anything, try to throw it off. It replies in Russian. The other addresses are about my path before agent work.
- Demo bot “Watch Officer” The same agent as in “live agents”. The data is invented — break it freely. t.me/vahtenny_demo_bot
- My professional track record A separate site about my product work: what I did, what I built, with which numbers. dervlad.comLinkedIn
- Message me directly I answer myself, within the day. No enquiry form and no account manager in between. t.me/der_vlad
Honest answers
What if the number doesn’t grow?
You pay nothing. The before/after review and the report stay with you — that’s worth something in itself: you’ll know your routines in numbers. I lose a week of work; that’s exactly why I review before starting and only take on cases where I can see the number really can be lifted.
Why a percentage rather than a fixed price?
So that I earn from the same thing you do. A fixed price pays the contractor even when the result is zero — a share of the uplift doesn’t. And the three-month term, agreed in writing, protects you from a percentage that never ends: after that the system is yours and the payments stop.
Can any routine be handed to an agent?
No, not any. The routines that automate well are the ones that repeat daily and live in messages, calls or spreadsheets: bookings, reminders, payment tracking, checking data. Write to me and I’ll tell you honestly whether I’ll take it on. I do say no: if AI won’t solve your problem, you’ll hear that on the first call, free.
Will the agent act without checking with me?
The agent doesn’t touch money at all: no payments, no price changes, no confirming refunds — not without a human approving it. A question outside the agreed scenarios isn’t guessed at; it is passed to you or to your office, flagged. What the agent said, to whom and when — you can open in the log at any moment.
How long does setting it up take?
A week: a day for the review, a day to measure the starting point, two days of building, a day to agree the scenarios and the limits — and launch on live enquiries by the end of the first week. Then a second measurement after the agreed period.
What happens to my data?
The agent and its logs run in an isolated environment in the EU. If we part ways, I hand them over to you or delete them — your choice. Business data is never sold or published. Only the necessary minimum reaches the language model — on API terms, with no model training on your data.
What of everything built stays with me?
All of it: the code, the bot, the knowledge base and the logs. Even if we part ways. Access is revoked in a minute — one key, you switch it off yourself and the agent stops immediately.
What you get from the free review
30–40 minutes · no preparation needed · commits you to nothing · the numbers are yours to keep
- We’ll find routine no. 1
I’ll go through the processes and show which piece of repeating work costs the most — in hours and in money.
- I’ll measure it as a mystery shopper
Where it helps, I’ll check myself, in the evening and at the weekend: how many enquiries go unanswered, and for how long.
- We’ll put the starting point in writing
So there’s nothing to argue about later. The threshold, the share and the term — before work starts, not after.
- I’ll say “no” if the task isn’t one for AI
This is a working meeting, not a sales call. If an agent won’t solve your problem, you’ll hear that on the first call, free.
Your free review
Find out your number — it’s free and commits you to nothing
Message me the word “review” — I’ll go through your repeating work, count what each one costs in hours and money, and send the numbers back in writing. What you do next is up to you.
Free review: where routine work is costing you