# About me, for AI assistants > Architecture and development for business systems, CRM integrations and customer applications. Get an initial estimate from your brief. This page is put together from the published texts of this site: who I am, what I do, how I work and where the details are. Nothing new is written here; the figures and terms are the same as on the home page. ## Who I am Complex business. Simple systems. I help businesses work more efficiently: automate routine work, connect data and introduce new technology where it pays off. - **I ran my own businesses** I used to work in sales: automotive from 2008, construction from 2014, and in 2015 I sold coffee. That same year I opened my own business, a marketing agency: I set up advertising, packaged franchises and built CRM and automation on no-code tools. Later I ran shops selling height-adjustable children's furniture. I started programming in 2017, when my companies needed systems nobody sold off the shelf. Now I build those systems for clients. It is my life's work, and I truly enjoy it. - **I start with your process** You show me how the work runs today. I find out who is responsible for what, where the data lives and at which step things get stuck. I dig deep here, because it decides what needs building and what can stay as it is. - **I raise constraints early** Before work starts, I separate what we already know about the task from what still needs testing. If your chosen approach won't work, I'll tell you and explain why. We agree on the scope and price of the first stage before development begins. - **I design the whole system** From the start I map how the screens, business rules, data and server connect. That way the task can be split into parts without losing sight of the whole. - **I hand routine work to AI** AI has made my development several times faster: it takes on routine work and helps with testing and debugging. I make the architecture and the main decisions myself, and I review what the tools produce. ## What I do What brings you here? - **Keep track of every enquiry** When enquiries arrive through calls, a website and messengers, it is hard to see who should respond next. A shared CRM gives the team one place to follow each request. I connect the enquiry sources to the CRM and set up the rules for assigning an owner, a status and the next action. - **Automate repeated steps** Sending the same reminder, copying a record or preparing another document takes time away from the work that needs a person. These repeated steps can follow agreed rules. I map the process and automate suitable reminders, data transfers and document preparation. - **See what your data is telling you** A spreadsheet or CRM can hold years of records without answering a simple question about workload or performance. A useful report starts with the decision you need to make. I bring the relevant data together, agree on how the figures are calculated and build reports or dashboards around those questions. - **Bring the workflow into one system** When bookings, files and customer details live in separate chats and spreadsheets, staff have to piece the order together. A connected website or internal app can carry those details through the process. I build the customer flow and the connections between the website, internal tools and business rules. - **Give AI a specific task** Reading documents or reviewing calls can involve a lot of repetitive work. AI is worth testing when the required output is clear and a person can check the result. I test recognition and extraction on representative examples, define what needs review and connect suitable steps to the workflow. - **Connect working equipment to your system** Equipment on different sites may need to exchange data and commands with one server. The first task is to understand the existing controllers and how they can communicate. I set up the network, server, database and connections between the controllers and the software that uses them. - [Equipment monitoring and operational alerts](https://kononov.dev/en/solutions/equipment-monitoring): Connect equipment readings, history and alerts in an interface that helps operators understand what needs attention. (Agriculture and production) - [Software and monitoring for food production](https://kononov.dev/en/solutions/food-production-automation): Connect production records, equipment measurements and operational handoffs without replacing specialist process decisions. (Agriculture and production) - [Greenhouse and hydroponics automation](https://kononov.dev/en/solutions/greenhouse-automation): Connect greenhouse controllers, climate readings and operational workflows with software designed around the installed equipment. (Agriculture and production) - [Irrigation and pump control](https://kononov.dev/en/solutions/irrigation-pump-control): Coordinate irrigation schedules, tank levels and pump states with explicit operating limits and manual control. (Agriculture and production) - [Monitoring and software for wineries](https://kononov.dev/en/solutions/winery-automation): Connect temperature records, cooling equipment and batch context in a practical workspace for winery operators. (Agriculture and production) - [AmoCRM integration with your website, spreadsheets and apps](https://kononov.dev/en/solutions/amocrm-integration): I connect AmoCRM (Kommo outside the CIS) to websites, Google Sheets, apps and your own server. Deals stay in AmoCRM, complex logic runs alongside. (Service and sales, Integrations and infrastructure) - [Bitrix24 integration with your own backend and tools](https://kononov.dev/en/solutions/bitrix24-integration): I build a separate backend or internal tool and connect it to Bitrix24. I also help you decide what stays in Bitrix24 and what moves out. (Service and sales, Integrations and infrastructure) - [CRM connected to the work behind it](https://kononov.dev/en/solutions/crm-integrations): Connect CRM records, incoming requests and team tasks with clear ownership and recoverable handoffs. (Service and sales) - [Workflow automation from request to next action](https://kononov.dev/en/solutions/workflow-automation): Team workflows with explicit states, responsibilities, exceptions and a useful first automation stage. (Service and sales, Data, documents and AI) - [Chatbots that hand work to the right system](https://kononov.dev/en/solutions/chatbots): Telegram, WhatsApp, Google Chat and web chat integrated with CRM and operational workflows. (Telephony and messaging, Service and sales) - [IP telephony connected to business operations](https://kononov.dev/en/solutions/ip-telephony): Asterisk, FreePBX and 3CX: routing, IVR, CRM connections, reporting and operational support. (Telephony and messaging, Service and sales) - [B2B portals for orders and customer service](https://kononov.dev/en/solutions/b2b-portals): Help business customers specify a request, approve documents and repeat orders through a portal connected to existing systems. (Customer products, Service and sales) - [A custom app built around a customer journey](https://kononov.dev/en/solutions/custom-apps): Customer applications and portals with an operational handoff and explicit first-release boundaries. (Customer products) - [Customer apps and portals connected to your operations](https://kononov.dev/en/solutions/customer-products): Customer applications and portals connected to the operations behind them. (Customer products) - [AI workflows with a result people can verify](https://kononov.dev/en/solutions/ai-workflows): AI assistance for defined document and operational tasks, with human review, visible uncertainty and spending controls. (Data, documents and AI) - [Google Workspace as a connected working environment](https://kononov.dev/en/solutions/google-workspace): Automate Sheets, Docs, Drive, Gmail, Forms and Calendar with Apps Script and external API integrations. (Data, documents and AI, Integrations and infrastructure) - [Python for documents, data and repeatable processing](https://kononov.dev/en/solutions/python-data-processing): Data analysis, OCR, API pipelines, permitted web data collection, visualisation and automation. (Data, documents and AI) - [Spreadsheet automation and business reporting](https://kononov.dev/en/solutions/spreadsheet-automation): Google Sheets and Excel: reliable data transformations, formulas, reports and integrations. (Data, documents and AI) - [1C integration: your own tools next to your accounting](https://kononov.dev/en/solutions/1c-integration): I connect 1C, the accounting and ERP system common in CIS countries, to websites, spreadsheets, documents and custom tools. 1C stays in place. (Integrations and infrastructure, Data, documents and AI) - [API integrations with explicit delivery rules](https://kononov.dev/en/solutions/api-integrations): Connect applications through data contracts, safe retries, visible partial failures and appropriate access boundaries. (Integrations and infrastructure) - [Infrastructure you can release and recover](https://kononov.dev/en/solutions/devops-infrastructure): Reproducible application builds, release checks, backups and a recovery process the team can follow. (Integrations and infrastructure) - [Custom internal systems for business workflows](https://kononov.dev/en/solutions/internal-systems): Custom business systems, CRM integrations and workflows designed around the way your team works. (Integrations and infrastructure, Service and sales) ## Where and in which languages I work How do we work across time zones? My schedule is flexible and arranged around clients in the USA, the CIS and Europe. I run projects in Russian and English. Site languages: English, Русский, ქართული, Türkçe. ## How I work From your task to the first launch For larger projects, I assemble a team to suit the task. I am responsible for the architecture and the result. You speak with me, without a chain of managers. - **You describe the task** Start with a brief or a conversation. Tell me how you work today and what you want to change. The task description and details in the brief are optional: you can leave just your contact details. - **We choose where to start** In a free discussion, I ask about the current process and its constraints. Together we identify a useful first stage and the result it should deliver. - **We agree on scope and cost** Before development begins, we agree on what the first stage includes and what it costs. Payment for that stage is 50/50: half before work begins, half after delivery. - **We check and launch** I prepare the system for launch; you check the agreed working scenario. We discuss further development and support separately. - **How is the work paid for?** We start with 50/50: half before work begins, half after the stage is delivered. Further work is agreed separately, under our agreements and defined tasks. - **Why now?** AI helps me with routine work, testing and debugging. I still make the architecture and key decisions, and the tools’ output needs checking. We agree on timing and cost based on the task and scope of the first stage. - **Who owns the code and server?** You do. The code, database, server and all access belong to the customer. Another developer can continue working on the system. - **What if you become unavailable?** The system keeps working on your server. The code is in your repository, and documentation and access are always handed over, so another developer can take over. - **Can you work with my current CRM?** Yes, with most popular systems: AmoCRM, Bitrix24, HubSpot, Pipedrive, GoHighLevel, 1C and others. If your CRM works for you, I connect your website, telephony, messengers and documents. If it no longer meets your needs, I can help you switch or build your own. - **Can we start small?** Yes, and that is often the best route. The first stage can be small, such as a $1200 calculator. You see the result and decide where to go next. That is how our work with a very large holding began. ## What it costs: examples - **How much does it cost?** It depends on the task. A small task, about $1200: an oil transport cost and volume calculator for a very large holding. It took 12 to 14 hours, covering design, calculations, 3 to 4 revisions, a server and a subdomain. A medium project, $5000 to $7000 over 1 to 2 months: CRM and telephony integration, from architecture through setup to launch. A larger project, about $14 000 to beta: an app for tracking oil deliveries and analysing discrepancies for the same holding, deployed on the company’s servers under its security requirements. One-off changes cost about $80 per hour. Ongoing support has a separate agreement. Send a sufficiently detailed brief for an initial cost estimate. ## Projects Projects and results - [Technical partner to a US PDR company since 2021: server, telephony, super app and document parsers](https://kononov.dev/en/cases/pdr-technical-partner): Since 2021 I've run the server, AI call analysis, super app and document parsers for a US PDR company. Lead flow multiplies yearly with almost the same team. (Service operations · Debian, Google Cloud, MariaDB, Node.js, Next.js) - [A service team lifted conversion by 30%: every call goes into the CRM, and genuine leads go through the full funnel](https://kononov.dev/en/cases/service-operations): Since 2021: backend, a 15-metric team leaderboard, real-time TV dashboards, GoHighLevel leads from IVR calls and AI call analysis. (Service operations · GoHighLevel, API, IVR, AI, Real-time dashboards) - [Full-stack development for a London consultancy, 2023–2024](https://kononov.dev/en/cases/consulting-partnership): Development for a London consultancy in 2023–2024: financial data, news scraping, quote capture, freelancer scheduling and Whisper subtitles. (Data, scraping and AI · Python, Node.js, Google Cloud, Debian, Morningstar API) - [A South African car rental hands over cars 3x faster: booking, documents and inspections linked to AmoCRM](https://kononov.dev/en/cases/car-rental-platform): How an AmoCRM integration grew into a system for a car rental business: full booking cycle on the website, calendar logic, documents, inspections and servicing. (Rental operations system · AmoCRM, Google Sheets, ReactJS, FlutterFlow, Firebase) - [Order calculation in 65 seconds instead of 4–5 hours: parts and materials from Excel BOMs, checked against 1C](https://kononov.dev/en/cases/manufacturing-orders): Software that counts parts and materials from Excel BOMs, checks them against 1C, consolidates orders and prepares work for shops and contractors. (Manufacturing · JavaScript, Node.js, Express, Electron, Tailwind CSS) - [AmoCRM, chatbot and messaging: a project outline](https://kononov.dev/en/cases/amocrm-chatbot): A CRM-centred project outline for Telegram, WhatsApp, web chat, telephony and operational reporting. (CRM & conversation · AmoCRM, Node.js, Python, Telegram / WhatsApp, Asterisk) - [The team saves over 100 hours a month: the service builds the group coffee order and fills the supplier's cart](https://kononov.dev/en/cases/coffee-group-orders): An internal service for pooled office coffee orders: wholesale catalogue, one supplier order, automatic cart filling, payments, receiving and handover. (Internal system · React, TypeScript, Node.js, SQLite, Docker) - [A custom API for connected applications: requirements and architecture](https://kononov.dev/en/cases/custom-api): Node.js, database models, permissions, documentation and integration contracts in an API project outline. (Backend architecture · Node.js, Express / Fastify, MariaDB, Redis, JWT) - [Debian on Google Cloud: an application infrastructure outline](https://kononov.dev/en/cases/debian-cloud): Compute Engine, web services, databases, release processes, monitoring and recovery in a cloud infrastructure outline. (Cloud infrastructure · Debian, Google Cloud, Nginx / Apache, MariaDB, systemd / PM2) - [A US PDR company saves about 60 hours a week: parsers read its insurance estimates, certificates and VIN photos](https://kononov.dev/en/cases/document-ocr): Python OCR parsers with AI validation for insurance estimates, work certificates and VIN photos. About 99.9% of documents go through without manual fixes, and a person only does the final check. (Documents & data · Python, OpenCV, Tesseract, Camelot, Pandas) - [Finance data, news API and AI chatbot in Google Sheets, 2023–2024](https://kononov.dev/en/cases/finance-data-ai-chatbot): Company data pulled into Google Sheets from finance sites or APIs, in an agreed format, with docs for the team and code protection. Upwork, 2023–2024, 5.0. (Financial data & AI · Google Sheets, Apps Script, REST API, AI, Data analysis) - [Clash-free freelancer meeting booking on Google Sheets and Calendar, 2023](https://kononov.dev/en/cases/freelancer-availability): A system for a consulting company: freelancers matched from Google Sheets data, availability from Google Calendar and meetings booked without clashes. (Scheduling & booking · Node.js, Google Sheets, Google Calendar, OAuth) - [GoHighLevel: a plan for pipelines, lead capture and communications](https://kononov.dev/en/cases/gohighlevel-crm): Sales pipelines, lead capture, communications, appointments and reporting in a GoHighLevel project outline. (CRM & sales · GoHighLevel, Twilio, Google Calendar, REST API) - [Google Sheets and Apps Script work from Upwork, 2021–2025](https://kononov.dev/en/cases/google-sheets-automation): Google Sheets scripts built for Upwork clients in 2021–2025: contact sync, sheet matching, emails, Looker Studio reports, Amazon S3 and MySQL. (Google Sheets & Apps Script · Google Sheets, Apps Script, Google Workspace, Looker Studio, REST API) - [A US hydroponic farm: 25 controllers on 2 sites in one network with a server and database](https://kononov.dev/en/cases/greenhouse-controls): How I linked a US hydroponic farm's controllers on several sites to a Google Cloud server with VPN, reverse proxies and MariaDB, supporting it since 2023. (Hydroponic farm infrastructure · Google Cloud, Debian, MariaDB, VPN, Reverse proxy) - [Inventory in Google Sheets with workflow automation](https://kononov.dev/en/cases/inventory-sheets): Stock movements, reorder alerts, barcode input, role permissions and management reports. (Inventory operations · Google Sheets, Apps Script, Google Forms, Mailgun / Twilio) - [Loan call center: an outline for connecting telephony to CRM](https://kononov.dev/en/cases/loans-call-centre): PBX routing, operator context, recordings, reporting and follow-up in a loan-service call-centre outline. (Telephony & contact centre · Asterisk / 3CX, SIP, IVR, CRM, Twilio) - [Event-driven email with Mailgun and Apps Script: a project outline](https://kononov.dev/en/cases/mailgun-automation): Personalised messages, spreadsheet data, Forms and Calendar triggers, delivery reporting and error handling. (Email automation · Mailgun, Google Apps Script, Sheets / Forms, Google Calendar) - [Mailwizz: email bounces and a domain summary in a 26-column sheet](https://kononov.dev/en/cases/mailwizz-analytics): A completed integration linking email bounces, Google Sheets, registration logs and blacklist rules. (Email operations · Python, MailWizz API, Google Sheets, CSV) - [A London consultancy collected news from 120 sources in one Python Scraper Hub](https://kononov.dev/en/cases/news-scraper-hub): A Python and Node.js service that collected news and materials from 120 sources, in full or by block, and showed each run's stage in real time. (News & data scraping · Python, Node.js, WebSockets, Nginx, Debian) - [An employee onboarding app: requirements and outline](https://kononov.dev/en/cases/onboarding-app): Roles, documents, training and reminders in an onboarding application with Google Workspace integrations. (HR & onboarding · Next.js, TypeScript, Tailwind CSS, Node.js, MariaDB) - [Python script: API data processed to a spec, December 2022](https://kononov.dev/en/cases/python-api-data-script): A Python script for a Liverpool client that downloads data through an API and processes it to a detailed requirements document. Upwork, Dec 2022, 5.0. (Python & APIs · Python, REST API, CSV) - [Quote Finder, 2023–2024: a quote highlighted in its source article, with screenshots and video](https://kononov.dev/en/cases/quote-finder): A tool that took an article link, found the quote, highlighted it in yellow in the original and saved screenshots and video of it to Google Drive. (Web page capture · Puppeteer, Headless Chrome, Google Drive) - [Referral service: a plan for rewards, CRM and payouts](https://kononov.dev/en/cases/referral-service): Referral links, conversion tracking, reward management and CRM/payment integrations. (Referral product · Next.js, Node.js, MariaDB, Stripe / PayPal, Twilio / Mailgun) - [SRT subtitles from video with a local Whisper model for bloggers, 2023–2024](https://kononov.dev/en/cases/video-subtitles-pipeline): A system for travellers and bloggers: Google Drive videos and podcast or YouTube links became SRT subtitles via a local Whisper model. (Video transcription & AI · Python, Flask, Whisper, MoviePy, Google Drive) - [Daily WhatConverts lead report in Google Sheets, 2022](https://kononov.dev/en/cases/whatconverts-reporting): An Apps Script that pulls lead counts from WhatConverts every day and totals them by source, type, week and month. Upwork, 2022, rated 5.0. (Lead reporting · WhatConverts API, Google Sheets, Apps Script) ## Lab - [Catch the requests](https://kononov.dev/en/lab/sort-requests) - [Team dashboard](https://kononov.dev/en/lab/tv-dashboard) - [From document to CRM](https://kononov.dev/en/lab/document-parser) - [Car booking app](https://kononov.dev/en/lab/booking-app) - [Game: a tomato greenhouse](https://kononov.dev/en/lab/greenhouse-game) - [Calculator: lost requests](https://kononov.dev/en/lab/roi-sales) - [Calculator: production losses](https://kononov.dev/en/lab/roi-agro) - [Calculator: manual data work](https://kononov.dev/en/lab/roi-docs) - [Calculator: abandoned orders](https://kononov.dev/en/lab/roi-product) - [Example: from request to work](https://kononov.dev/en/lab/request-flow) - [Brief with a preliminary estimate](https://kononov.dev/en/lab/project-brief) ## Insights - [AI in data processing: extraction and review](https://kononov.dev/en/insights/ai-in-data-processing): Where AI helps interpret documents and messages, how to validate its output, and when a person should decide. - [Background workers: jobs, queues and retries](https://kononov.dev/en/insights/background-workers): Run longer tasks outside the user request, with visible queue states, controlled retries and duplicate handling. - [Business logic: rules, states and permissions](https://kononov.dev/en/insights/business-logic-and-rules): Define validation and workflow transitions so AI, people and background jobs follow agreed rules. - [What to decide before building a car rental app](https://kononov.dev/en/insights/car-rental-app-planning): Availability, booking ownership, documents and exceptions as design questions. - [Planning the first release of a customer portal](https://kononov.dev/en/insights/customer-portal-first-release): Start with roles, ownership and one complete customer journey. - [Connecting documents, CRM and tasks reliably](https://kononov.dev/en/insights/reliable-crm-workflows): Design for duplicates, missing information and failed handoffs before adding AI. - [Remote desktop on Debian 11: the original setup](https://kononov.dev/en/insights/remote-desktop-debian): How the original Debian 11 and Google Cloud setup used LXDE and XRDP, and what to review before a new deployment. - [The server: data, access and workflow coordination](https://kononov.dev/en/insights/server-as-system-core): Why a system needs a shared core that preserves state, checks permissions and coordinates data processing. - [When is it time to replace spreadsheets?](https://kononov.dev/en/insights/spreadsheets-to-systems): Five practical signals, and a way to scope a useful first system. ## Contact Let’s discuss your task Start with your name and email. Add a task description and further details if you’d like. - [Discuss your project](https://kononov.dev/en/contact) - [Email](mailto:dmitrii.kononov.ge@gmail.com) - [Telegram](https://t.me/Dimetrrius) - [WhatsApp](https://wa.me/79124631686) - [Upwork](https://www.upwork.com/freelancers/dimetrius?mp_source=share) - [LinkedIn](https://www.linkedin.com/in/kononov-dmitriy) ## Other languages - [Русский](https://kononov.dev/ru/ai) - [ქართული](https://kononov.dev/ka/ai) - [Türkçe](https://kononov.dev/tr/ai) --- I did not write anything new for this page: everything comes from the published pages of this site. Figures and terms appear as published on the home page. [Business systems and customer products](https://kononov.dev/en) · [llms.txt](https://kononov.dev/en/llms.txt)