#Hiring right now is a statement in itself. In 2026, when AI is cutting into headcount everywhere, we're doing the opposite - growing our fantastic team! We're looking for someone who can sell complex technical work: - talk to technical buyers as an equal, run a long sales cycle, and close cold leads without hand-holding. Reach out or tag someone who fits: https://lnkd.in/dmNgaAUw #vacancy
INC4
IT Services and IT Consulting
Kiev, Kiev 725 followers
AI and blockchain engineering studio 🚀
About us
INC4 is an engineering studio founded in 2013, headquartered in Kyiv with an office in Lisbon. 70+ certified engineers across five practices: AI Lab (LLMs and autonomous agents), Blockchain Hub (smart contracts and DeFi), Algotrading (execution engines and HFT infrastructure), MLOps & DevOps, and Compute Infrastructure (bare-metal GPU colocation). Partners include Nvidia Accelerator, AWS, and NEAR. Trading systems managed: $10B+ monthly Our studio operates across five specialized practices: · AI Lab — Custom LLMs, multi-agent orchestration, neural pipelines, and autonomous AI systems for finance and blockchain. Our engineers contribute to LangChain AI open source and publish research on AI agent safety, including Survivability-Aware Execution (SAE) for live trading agents and the Interaction-Native Knowledge Harness (InKH) for financial LLM memory architecture. · Blockchain Hub — L1/L2 infrastructure, DeFi protocols, smart contracts, DAO systems, and crypto wallets. 100+ on-chain projects delivered since 2013, including AirDAO (full Layer 1 blockchain) and Open Forest Protocol. · Algotrading — Backtesting infrastructure, OMS/EMS, signal processing, and quant research platforms. Built by a team that runs on its own capital. · MLOps & DevOps — High-load cloud-native architecture, CI/CD pipelines, and infrastructure automation. AWS ML, Kubernetes (CKA/CKAD), and HashiCorp Vault certified engineers. · Compute Infrastructure — Bare-metal, colocation, and edge environments for workloads that can't live in the cloud. We work as an embedded engineering partner, not an outsourcing body. Our teams integrate with yours — senior engineers (10+ years average experience), certified across AWS, Kubernetes, and HashiCorp, who ship production systems and contribute back to the ecosystem.
- Website
-
http://inc4.net/
External link for INC4
- Industry
- IT Services and IT Consulting
- Company size
- 51-200 employees
- Headquarters
- Kiev, Kiev
- Type
- Privately Held
- Founded
- 2013
- Specialties
- blockchain and development
Locations
-
Primary
Get directions
Kiev, Kiev 0, UA
-
Get directions
Tallin, EE
Employees at INC4
Updates
-
Today Ukraine turns 35🩵💛 It's more than a date on the calendar - it's a reminder of why we do what we do. Thank you to everyone defending this chance to live, create, and build. #Happy_Independence_Day! #2026
-
-
#Monaco is where some conversations stop being theoretical. At WAIB Summit Monaco, our Head of Strategy & Ventures, Julia Stadnyk, led a discussion on one of the defining questions for the next era of finance: - Who wins, who gets left behind, and what happens when Bitcoin, digital assets, AI and regulation start shaping the same market at the same time? This is the kind of conversation we believe matters, because the infrastructure behind it is being rewritten right now. Access, trust, compliance, intelligence, execution, user experience - all of these layers are starting to converge. And the companies that understand this shift early will not simply adapt to the future of finance. They will help define it. Proud to see Julia representing INC4 on stage in Monaco and bringing our perspective into one of the industry’s most important conversations 🔥
-
-
Another big step unlocked! Our latest paper is live - and it captures one of the questions we keep coming back to in our AI work: - what should an agent remember, and what should it forget? The idea is simple but powerful: - better AI agents don’t need to remember everything. They need to know what to forget. https://lnkd.in/dXG-4hua
Everyone's racing to give AI more memory. We went the other way. Most agents fail not because they forget but because they remember the wrong things: - yesterday's market sentiment - last week's news - stale context that made sense 3 months ago but just kills decisions today... So we built a framework around governed forgetting. Three ideas behind it: 1. The system feeds context to the agent, not the other way around 2. Outdated knowledge gets dropped the moment new evidence contradicts it 3. Аgent works with what matters now, not its entire history 📌 46,000 evaluations. 83% lower latency. 97% fewer stale answers. Big thanks to Ailiya Borjigin and both True Trading & INC4 teams for making this real 🔥
-
Next week, our INC4 leadership will be in #Monaco🔥 We’ll be there for WAIB Summit Monaco week - from the F1 Paddock experience to conversations with builders, founders, investors, and teams working at the intersection of AI, crypto, finance, and product strategy. Julia Stadnyk will also take the WAIB Summit Monaco stage to talk about how crypto and Bitcoin are rewriting finance. If Monaco is on your map next week, come say hello😊
-
-
Our Max Chikita was at Amazon Web Services (AWS) Summit Warsaw 2026 - diving into sessions on Symbolic AI, Knowledge Graphs, and agentic systems... 80+ talks, hands-on labs, and the full cloud frontier in one day 🚀 That’s the kind of energy we bring to INC4. Mre to come! #AWSSummit #AWS #AI
-
-
AI is moving from “help me think” to “do this for me.” That shift is HUGE. But the more AI can act, the more important it becomes to control what it is allowed to touch. Our CEO Igor Stadnyk shared more on this in his latest piece for CCN - Crypto Citizens Network, esp. through the lens of AI trading and agentic finance. https://lnkd.in/dFrTDfWq
The dangerous part of AI is not that it can think. It is that it can ACT. A bad summary wastes time. A bad action can move money, trigger workflows, open positions, or create losses nobody meant to create. What still surprises me is this: - Before, we used to think twice before accepting cookies. Now we give AI access to files, tools, accounts, credentials, sometimes even the whole computer easily - often just to solve what feels like a small task. That is where the risk starts. In many cases, you think you delegated one simple thing. In reality, you may have given the system much more permission than you realize. So... Smarter AI matters. But smarter boundaries matter even more Shared more on this in my latest piece for CCN - Crypto Citizens Network. Link in the first comment.
-
-
EXCLUSIVE 🔥 Where is AI actually going? Our CEO Igor Stadnyk shares his view in the interview below👇
AI won’t beat top quant firms. And that’s not even the point. Firms like Jane Street or Two Sigma won’t be outperformed by AI, not on signal quality & not on execution speed. Because they have structural advantages: - infrastructure, capital, data access, and decades of optimization... You can’t close that gap by just adding AI on top. So the real question is different: - where does AI actually add value? Shared some thoughts on this in a recent piece 👇 (link in the first comment)
-
-
Safe execution will be a key layer for AI agents. If you haven’t heard about #SAE yet - check out the research by our CEO Igor Stadnyk and team.
AI is moving very fast, and AI agents are already shifting from analysis to execution. That changes the problem. The main risk is no longer only model quality. It is SAFE execution. So we introduced the first version of #SAE (Survivability-Aware Execution) - a framework for safer AI agent execution under real-world conditions. Early results already show the approach works. Paper is now out - link in the first comment.
-
