BLACKBENCH

5 projects · 2 people

Built with AI.
With judgment
and real experience.

Websites, product prototypes, AI agents, trading systems, internal CRMs and information systems of any complexity.

Write to us

years of practice15+
screens designed85
systems in flight5

We are a team with more than 15 years in IT, across web and fintech. Working with AI speeds development up sharply, and that is what lets us ship quickly and price sensibly.

How we work and how you pay

Landing pages and websites

  1. We build the whole thing and host it on our servers.
  2. You pay 100% of the fee.
  3. We move it to your infrastructure and set it up.
  4. Two weeks of support at no charge — fixes and small adjustments. Critical or large changes are agreed separately.

AI agents and complex information systems

  1. We produce the design, an interface prototype and a detailed spec.
  2. You pay 20% of the total.
  3. We build exactly what the spec agreed.
  4. You pay 30%.
  5. Trial operation for one to two months, as agreed: the system runs on real work while we make fixes and small adjustments. Critical or large changes are agreed separately.
  6. You pay the remaining 50%.
  7. Production use plus a month of support at no charge: we fix bugs and make small adjustments. After that, paid support and larger changes by agreement.

Team

Developer and team lead · 20+ years

Boris Kraportov

Briefing, architecture and AI-assisted development.

QA Lead · 15 years

Ruslan Zeynalov

Quality assurance: manual and automated testing.

Pricing

Websites and landing pagesfrom $350
Websites with a CMS (an editor for adding pages and blog posts yourself)from $450
AI bots on our RagReply platformfrom $200/mo*
Custom AI agents, chatbots and similar buildsfrom $750
Custom information systems at any complexity — internal CRMs, financial accounting, non-trivial AI agents, plus integration with outside platformsfrom $8,500
Trading bots, PineScript/MetaTrader scripts and integrationscustom quote

Starting points, not fixed quotes: one-off work is priced per project, the RagReply platform monthly. Exact quote follows the brief.

* Model tokens are billed separately when demand runs high.

Work

Design & product · 2026

iqplay

A platform for club tennis with one shared rating and several tournament formats.

Amateur players get one measure of ability they can all agree on, and a tournament that fills up and is over by the evening.

Every screen in one place — 35 of them, grouped by flow, from NTRP self-assessment at sign-up through to the admin panel.

Design & product · 2026

aura

A medical CRM with AI-driven booking and a patient record built to be used.

One system for the schedule, the medical record, messaging-app booking and marketing — instead of four that never quite agree.

Every screen in one place — 45 of them across four portals, ordered by scenario from patient sign-in through to AI settings.

Trading · Solana · 2026

pumpgod

A collector and sniper for pump.fun token launches, riding a Yellowstone gRPC stream.

Your own record of a market's first minutes, so a decision never waits on somebody else's screener to catch up.

The live feed of new tokens — buyers in the last 30 seconds, SOL in, peak reserves and clone count, one row each. Every screenshot in this case study comes from an isolated demo database of synthetic data, never the production archive.

AI systems · 2026

RagReply

An admin panel for AI agents that answer from the client's own knowledge base.

Upload the documents a business already keeps, and the agent answers by text, by voice or on a call — naming the file it took the answer from.

Clients, each with a knowledge base and a document count of its own. One panel runs several agents. The salon, the dental clinic and the garage were invented for this portfolio and seeded by seed_ragreply.py — they are not a client roster.

Trading · backtesting · 2026

polygangster

Backtesting a Smart Money Concepts order-block strategy against real Polymarket binary-option odds.

An honest answer for a trading hypothesis, taken from the odds that really stood in the book, before any money goes near it.

Setting up a run: instruments, timeframes, the structure set and the staking parameters, with finished sessions listed below. The win rate and the PnL in that log were computed over generated stakes from an isolated demo database: they measure nothing and promise nothing.

Get in touch

Tell us what needs building — we will come back with an estimate and a plan.

Write to us

hello@blackbench.dev
@blackbench