What does Veles Digital build?

Custom AI tools, automation pipelines, conversational chatbots, and analytics dashboards. We engineer practical AI systems that eliminate waste, accelerate growth, and ship. Not demos. Every project is shaped around the specific workflows of a client's business.

How long does a custom AI project take?

A typical first-version delivery is two to eight weeks. Discovery and specification take about a week, a focused MVP runs two to four weeks, and larger integrations or analytics dashboards extend to six to eight weeks. We ship working software in short cycles rather than long-lead waterfalls.

Do you work with non-technical clients?

Yes. Most of our clients are not engineers. We translate business problems into AI solutions, handle the technical complexity end-to-end, and deliver tools your team can actually use. Onboarding includes training and clear documentation.

What industries does Veles Digital serve?

We're industry-agnostic. The common thread across our clients is teams who want concrete AI-driven outcomes like automation, better analytics, and customer support efficiency, rather than experimental AI projects. Each engagement is shaped around your specific workflows and operational reality.

How does a project start?

Book a free 30-minute discovery call via calendly.com/velesdigitalsolutions/30min, message us on WhatsApp at +381 69 251 1196, or email contact@veles.digital. We discuss your business, identify which workflows AI can improve, and propose a scoped first project with a fixed price and timeline. We respond within 24 hours.

Where is Veles Digital based and who do you serve?

Veles Digital is headquartered in Serbia and serves clients across Europe and worldwide. We operate fully remotely with English and Serbian language support, and most engagements happen via async tooling and scheduled video calls.

How much does custom AI development cost?

We don't publish fixed prices because every project differs in scope, integrations, and timeline. The discovery call is free, runs 30 minutes, and ends with a scoped quote based on your specific needs. Typical engagements range from focused two-to-four-week MVPs to six-to-eight-week multi-system integrations, with fixed scope and fixed price for the agreed phase.

What's the difference between Veles Digital and a ChatGPT subscription?

ChatGPT is a general-purpose assistant; Veles Digital builds production AI shaped around your specific workflows, integrated with your data, deployed on your own stack. ChatGPT helps individuals draft emails. We build systems that automate your support inbox, qualify your inbound leads, surface insights from your sales calls, or pull together a board-ready dashboard from your real-time data. Owned by you, not rented per seat from OpenAI.

Which AI platforms and models do you build on?

We're model-agnostic. We work with OpenAI (GPT-4o, GPT-4.1), Anthropic Claude (Sonnet, Opus, Haiku), Google Gemini, Mistral, and open-source models (Llama, Qwen) depending on the use case. Selection is driven by cost, latency, accuracy, and data-residency requirements, not vendor loyalty. We also use vector databases (Qdrant, pgvector, Pinecone) and orchestration frameworks (LangChain, custom agentic patterns) where they make sense.

Can you integrate AI with our existing CRM, ERP, or data warehouse?

Yes. Integration is the core of every project. We connect AI capabilities to Salesforce, HubSpot, Pipedrive, Notion, Airtable, Postgres, MySQL, BigQuery, Snowflake, and most modern SaaS APIs. If your stack has a documented API or webhook surface, we can wire AI into it. We don't ask you to migrate your data to a new platform.

Do you provide ongoing AI maintenance and updates?

Yes. After delivery, we offer optional retainers covering production monitoring, model upgrades, prompt tuning, and feature evolution. Most clients move to retainer after the initial build to keep their AI current as models and APIs evolve. What worked on GPT-4o today may be obsolete or expensive six months from now, and a small ongoing engagement keeps the system competitive.

How do you measure AI ROI for our business?

Every project starts with a measurable target like hours saved per week, response time reduction, conversion lift, support ticket deflection, or qualified leads per month. We instrument the AI system to report on those metrics from day one. If we can't measure the win, we don't ship it. Discovery includes a baseline measurement so the impact is provable, not anecdotal.

// AI Agents · E-commerce

AI agents for ecommerce that close orders, not just route tickets.

Most "ecommerce chatbots" are flowchart tools that bounce a customer into a human when the question gets real. A custom AI agent owns the conversation end to end: pricing, stock, substitutions, order creation. Humans approve the unusual cases. Built around the channel where your buyers actually message you.

The problem ecommerce AI tools usually solve badly

Most ecommerce stores already have a "chatbot" of some kind. A scripted FAQ widget on the product page. A help-desk autoresponder in support email. A Shopify app that greets returning customers. They handle 30-50% of inbound messages and dump the rest on a human, often the owner.

The remaining 50-70% is where the actual revenue lives. A returning buyer asking about a product variation. A bulk order with custom shipping. A "do you stock X like the one I bought last month" question. A pricing negotiation. These are buying signals, and they get answered slowly or not at all because the tool that should help is structurally too dumb to understand them.

The cost of a slow reply is well-documented in B2C and B2B alike. Firms that contacted prospects within 5 minutes were 100 times more likely to make contact than those that waited 30 minutes (Oldroyd, McElheran & Elkington, Harvard Business Review). Every message that waits twenty minutes for a human is a basket your competitor closed first.

How we build it differently

We do not ship a chatbot. We ship a vertical AI sales agent built around your specific catalog, your pricing rules, your stock data, and your channel mix.

The agent reads every inbound message on whichever channel matters (WhatsApp, Instagram, email, web chat), understands the intent (price quote, order, support, off-topic), looks up the customer history, drafts a reply, and either sends it or queues it for owner approval depending on a small set of rules you define.

For repeat buyers on standard products at standard pricing, orders flow straight through. For new buyers, unusual quantities, or pricing edge cases, the agent escalates with a one-screen summary the owner can approve or edit in five seconds. The owner is no longer the inbox; they are the approval layer.

Tool use over pure prompting is non-negotiable. The agent does not "decide" prices, it calls a function that returns the real number. Same for stock checks. Auditable, deterministic, and impossible to hallucinate a discount that breaks margin.

What we ship for an ecommerce client

Conversation reader: agent classifies intent on every inbound message (price quote, order, support, refund, off-topic)
Customer memory: every buyer recognized by name, history, prior orders, and channel preference
Pricing function: pulls live margins, applies tier discounts, returns a deterministic number. No hallucinated quotes
Stock + substitution function: real-time inventory check with sensible alternatives when an SKU is out
Order creation: standard orders flow to your back-office without human review; edge cases escalate with full context
Owner dashboard: every active conversation visible in real time, one-tap takeover, one-tap return to the agent
Daily promo proposal: agent suggests which slow-moving SKUs to push to which buyers, owner approves in 5 minutes

// Proof from production

Already running in production

AI Commerce Stack

Mobilni Market

We built exactly this for Mobilni Market, a Serbian B2B mobile-parts wholesaler running hundreds of SKUs and dozens of recurring customers across the Balkans. The owner was personally answering 80-120 WhatsApp messages a day. Four weeks later the AI was reading every inbound, taking routine orders, proposing the daily promo, and escalating only the unusual cases. The owner went from 5 hours in WhatsApp to 30 minutes. The same architecture extends to any ecommerce operator whose volume has outgrown their inbox.

Case study

// FAQ

Frequently asked

Will this replace my Shopify / WooCommerce / Magento integration?

No. The AI agent sits alongside your existing ecommerce platform, not on top of it. We integrate via your platform's API or order endpoints. The agent creates orders, the platform processes them, your existing checkout / payment / fulfillment flow stays unchanged. If you switch platforms in two years, the agent moves with you.

What channels can the agent operate on?

WhatsApp Business API is the most common for SMB / B2B (we use the official Meta Cloud API, not third-party scrapers). Instagram DMs at scale, web chat embedded in your store, and email are all standard. The agent handles multi-channel conversations as one identity per customer, so a buyer who pings on WhatsApp and follows up by email gets a coherent response.

How do you stop the agent from giving away the store on pricing?

The agent never sees the pricing logic directly. It calls a pricing function we wire to your real margin rules. The function returns a number, the agent quotes that number, the conversation is logged for audit. Anything outside the approved pricing range (custom quotes, large-volume deals, irregular SKUs) escalates to you for approval before the agent commits.

How long until I see results?

A focused single-channel ecommerce AI agent ships in 3-5 weeks from kickoff to production, depending on integration complexity (one ecommerce platform + WhatsApp is the fast path; multi-platform + custom CRM is the slow path). The full scoping and milestone breakdown lives on our process page. The first 1-2 weeks post-launch are pure observation: we watch how real buyers interact, what the agent gets right, what it gets wrong, and tune. By month three the agent is handling the majority of inbound traffic at the quality bar you set.

What does this cost compared to a chatbot platform like Intercom or ManyChat?

Different category, and we dig into the actual decision tree in AI agent vs chatbot. Chatbot platforms charge per-seat per-month and cap your conversations or features. A custom AI agent is a fixed-scope project. The cost is paid once on build, plus a predictable monthly for monitoring and tuning. The break-even versus a per-seat tool is typically 12-18 months, and the agent handles a meaningfully higher share of inbound (85-95% vs ~50-70% for flowchart bots).

Do you work with non-Serbian ecommerce companies?

Yes. The AI layer is language-agnostic. Frontier language models handle most European languages natively, including code-switching between English and the local language mid-message. We have shipped multilingual systems for Balkan + EU clients. The integration work and back-office mapping is the same regardless of language.

How we estimate the work

Pricing follows scope, not a fixed rate card. The full breakdown of how we scope, ship in milestones, and what each phase includes lives on the process page.

See our process →

Ready to talk about a e-commerce project?

A discovery call is free and runs about 30 minutes. We map the problem, tell you if we are the right fit, and walk out with a first-milestone outline.

book a call Email us directly