Dmitrii Maslennikov

CTO of a group running its own products in real estate, logistics and fintech. I set the technical strategy, choose the stack for each product, build and lead the teams, and stay hands-on in the parts that decide whether a product survives contact with production.

AI is the through-line of the portfolio: LLM and RAG services that read unstructured business data - contracts, listings, support threads, market feeds - and turn it into decisions people act on.

• Built the retrieval, evaluation and cost-control layers that keep LLM inference affordable and stable at production volume, rather than impressive in a demo.
• Bayside Residence and the sister developments (Woodside, Lakeside, Highwood): a full operating stack for country estates outside Moscow - resident applications, service billing and payment, accounting, support desk, access control and security, food delivery, and generated feeds for advertising systems. One codebase runs several developments.
• Vuvo App - freight transportation in the United States, an Uber-shaped product for cargo. Delivered the backend, the Android and iOS applications, the administration panel and the public site.
• Integrated Stripe for trip payment and automatic payouts to drivers - the part of a marketplace that has to be right every single time.
• Hire, mentor and run the engineering teams across the portfolio, and own delivery end to end.

Lead Developer across the group's fintech and crypto products, from architecture through to release.

• Built ETF trading applications with an AI layer that analyses market data and produces signals investors act on - specific enough to be useful, conservative enough to trust.
• Carried that work into a fintech platform for a UAE bank, where AI reads market information and generates investor signals automatically; we went deliberately deeper on data analysis than the established players in that niche.
• Rebuilt the platform's interface around engagement, with gamification aimed at investors who return weekly rather than once a quarter.
• Delivered banking applications for Forte Bank.
• Shipped Qavala, a crypto trading platform: exchange, deposits and withdrawals, bank-grade security, direct blockchain integration and integration with the major exchanges - frontend and backend both mine.

Lead Developer on a token exchange built on the Solana stack.

• Delivered the gateways that swap the native token against USDT, ETH, BTC and other assets.
• Took wallet interaction from "mostly works" to reliable - the part of a crypto product where every edge case is somebody's money.
• Wrote the smart contracts in Rust, including staking and deposits, and the on-chain integration behind them.

Lead Developer on products where the hard part was regulation, money movement and biometrics.

• NaChai - money transfers by face recognition. You register by taking a photo of yourself, and from then on your biometrics are the account: you can send money to anyone, including a recipient who has never installed the app, with the amount held until they register and claim it.
• Built NaChai with a large Russian bank and a venture fund, originally for cashless tipping in restaurants; the pandemic closed that market before the product could take it.
• ESR Solutions - a crypto exchange platform for BTC, ETH, BCH and XRP plus the platform's own token, launched on Ethereum. Full cycle on my side, from prototyping and design through smart contracts to direct blockchain integration.
• Integrated Visa and Mastercard to issue physical cards linked to a user's account balance - where crypto stops being a wallet and becomes a payment product.
• kassa.taxi - cloud fiscalisation for taxi fleets under Russian 54-FZ: a correct receipt filed for every ride without each driver contracting a fiscal operator himself. Built the service, the internal CRM the fleets work in, and the application.
• Delivered goloft.ru.

Full Stack Developer at an e-commerce business selling professional cosmetics to salons and to consumers, and the engineer responsible for most of what the business ran on.

• Built the storefront and the catalogue behind it - a range where one product line carries dozens of shades and volumes, and where a wrong stock number is a refund rather than an inconvenience.
• Built order processing, payment, delivery and returns.
• Built the B2B side a professional range needs: separate pricing for salons and distributors, account-level discounts, and bulk ordering that behaves differently from the retail basket.
• Connected the shop to the warehouse and accounting systems, so stock, prices and orders stayed one truth instead of three.
• Owned search, filtering and merchandising - what decides whether a large catalogue sells anything at all.
• Six years on the same business, through several redesigns and one full replatform.

Full Stack Developer on the web presence of the haute joaillerie division.

• Delivered sites for the division to the standards of a maison where brand guidelines are not suggestions.
• Worked at a level of visual fidelity where a compression artefact on a stone is a visible defect, and the interface has to stay quiet enough for the pieces to carry the page.

Full Stack Developer at a full-cycle digital agency, delivering custom sites and applications for clients.

• A new business domain every few months, which is where a developer learns breadth.
• Direct contact with the client: scoping the work, and explaining the technical trade-off behind a decision to someone who does not think in those terms.
• Estimates that had to hold, against deadlines that did not move because the technology turned out to be harder than it looked - where I learned to defend an estimate and to say what a project would really cost before it started.

Alion is an AI hiring platform I designed and built - a search engine for engineering talent that reads what people have actually shipped instead of what their CV claims.

• Builds a verified profile from public work - GitHub repositories, papers, projects, public activity - and matches candidates to vacancies with explanations a hiring manager can check, not a similarity score they have to trust.
• Retrieval-augmented generation over the whole corpus, vector search across several engines, embedding pipelines, and an evaluation layer that keeps matching honest as the models and the data drift beneath it.
• Crawlers and adapters holding a live corpus of vacancies from hundreds of applicant tracking systems and job boards across dozens of countries.
• Salary estimation for the majority of postings that hide their pay, and detection of the postings that are not real jobs at all.
• Architecture, machine learning, infrastructure and production operation are all mine.

Noco is a developer of premium country real estate. For them I built the websites, the internal CRM systems that run the estates, and the apps for the residents.

• Internal CRM for managing the estates: electricity, bills, services, bookings and access passes.
• Android and iOS applications for the residents of the estates.
• Multiple marketing integrations.
• AI agents: a support assistant, and assistants for the marketing and sales teams.

jcvd.online, built with Jean-Claude Van Damme - bridging physical art, global pop culture and Web3 infrastructure, and exploring his legacy not as an action icon but as an artist working between traditional art and high technology.

• The physical half: an exclusive exhibition of his "Cities of the World" paintings at the University Museum and Art Gallery in Hong Kong.
• The digital half was ours - his personal token, $JCVD, launched on Solana.
• An e-commerce platform through which collectors anywhere buy his physical works and digital assets in one flow.
• Connecting an offline museum experience in Asia to Web3 mechanics was a logistical and engineering puzzle: custody, settlement, provenance and shipping all had to agree with each other, on a date fixed by a museum rather than by us. It shipped, on time, in Hong Kong.

Vuvo is an Uber-style platform for moving in the United States, with Android and iOS applications.

• A user orders a driver for a move or for transporting a car, and the platform finds the best-matched drivers.
• Automatic background checks of the drivers.
• An AI calculator that estimates the cost of a move from many parameters.
• An AI agent that accompanies the move.
• An admin panel for managing the platform, its users and its orders.

GoLoft is an online platform for booking lofts - an Airbnb made for lofts only.

• Loft search by parameters and on a map.
• A booking management system for the venues.
• A venue account with every setting a venue needs: photos, parameters and a Google Calendar connection.

ATR.ONE is an AI platform that finds freelance work for people in creative professions - models, actors, photographers, performers - and it runs on machine learning end to end.

• Collector robots read the popular social networks and the specialist sites; the AI layer analyses over 15 billion posts a day and pulls the genuine casting and booking calls out of that noise into one feed.
• The hard part is precision rather than volume: a creative professional abandons a feed that wastes their time twice.
• Works across four languages, focused on Russia, Kazakhstan and the UAE, with further markets planned.
• Real-time messaging over WebSocket, including group chats built to hold large numbers of participants.
• Built entirely in-house - design and prototyping, the code, the machine learning and the network architecture.

Rhinotales is a crypto game built on AI that trains game characters to move, made with ML-Agents in Unity on the Solana blockchain.

• Every player uses AI to train their own characters for crypto - fantastic beasts, ostriches, dinosaurs, dogs and more.
• Trained characters are entered into races: every entrant pays in, and the one that runs best wins the cash prize.
• Built on Solana.

NaChai is an app for sending money to a waiter by their face: open the app, take a photo of the waiter and send the money right away.

• The waiter receives the money by registering in the app and signing in with their face.
• An in-house face recognition engine built on AWS.
• An admin panel for managing the whole process.
• Integrations with a number of payment systems.

ESR Solutions is an international cryptocurrency exchange.

• Exchange, trading, deposits and loans.
• Support for all the popular tokens.
• Direct integration with the blockchains.