Info
Activiteit
3K volgers
-
Vadim Chepegin heeft dit gedeeldVadim Chepegin heeft dit gedeeldWorkshop voor MBO-studenten: AI in de praktijk Gisterochtend mocht ik een workshop geven aan de eerstejaars Handel & Ondernemen MBO-studenten van ROC Amsterdam. Ik vond het echt superleuk om te doen en was erg onder de indruk hoe snel de studenten EU GPT eigen maakten. Na bedrijfsonderzoek via AI websearch en eigen analyse waren er top presentaties op basis van AI-gegenereerde PowerPoint-slides. Deze studenten gaan het verschil maken in bedrijven. Door AI slim in te zetten, kunnen organisaties zich weer richten op hun kern. Zij voegen de waarde toe, zij bouwen de toekomst. Succes allemaal met jullie opleiding! Hayat E.iding! Hayat E., Kevin van de Kamp Bedankt voor deze mogelijkheid!
-
Vadim Chepegin heeft dit gedeeldRecently, my former colleague came to me with this article in hand and in frustration. A vacancy can look completely real. The company exists, the description is detailed, and the application form works. But the role may already be filled, frozen, unfunded, or posted without any immediate intention to hire. #ProofLine cannot guarantee that a job is real. I don’t think any candidate-side tool can. What it can do is make the search less blind and reduce the amount of time you lose when a vacancy goes nowhere. 1. Check the original company or ATS page and compare duplicate listings across boards 2. Keep the role, source, agency, dates and outcome in one place 3. Prioritise jobs by fit, evidence, location, pay and source quality 4. Reuse approved facts and answers instead of starting from zero 5. Learn which employers, agencies and repeated listings waste your time 6. Keep your data local, submit manually, and track what happened. This does not solve ghost jobs. That requires more transparency and accountability from employers, platforms and #Regulators. #ProofLine #SDG8 #DecentWork #FutureOfWork #JobSearch #CareerDevelopment #SDG4 #LifelongLearning #ResponsibleAI #HumanCenteredAI
-
Vadim Chepegin heeft dit gedeeldInteresting paper, with direct relevance to both #ProofLine and #EUGPT custom development work.Vadim Chepegin heeft dit gedeeldAnother interesting #IUI2026 session with several interesting papers: "AI-Augmented Work: Research and Education". Trying to post some that were especially interesting for me. The paper "From Toil to Thought: Designing for Strategic Exploration and Responsible AI in Systematic Literature Reviews" presented by Rulong Ye focused on human-AI collaboration to support writing of Systematic Literature Reviews, sometimes called PRISMA reviews. Many readers compain about PRISMA style reviews finding them shallow and incomplete, but the team suggested and explored several ideas to make these reviews both better and easier to write. Joint work with Naaz Sibia, Angela Zavaleta Bernuy, Tingting Zhu, Carolina Nobre, Viktoria Pammer-Schindler, and Michael Liut See the paper here: https://lnkd.in/gK_sXjqk
-
Vadim Chepegin heeft dit gedeeldWhat do you think: fair points or nuanced?Vadim Chepegin heeft dit gedeeldNo, I will not take your "equity" as payment. This is becoming a trend, especially across Italy. A founder has nothing but an idea, reaches out to professionals for the execution, and then, right when the quote lands, out comes the magical sentence. "I had something different in mind." LOL. Professionalism, you pay for it. Always. And yes, we do take equity sometimes, when we believe in the project, when the founder is serious, and when the plan is for us to stick around well past the build. But that is a structure, with a valuation cap, a discount, a SAFE, and a cash component that shows you have skin in this too. What it is not, is a number you invented on a Sunday afternoon and now expect me to accept as a salary. The worst part is the pitch that comes with it. The idea that you are doing me a favour, that this is somehow a better reward than money, that I should be grateful for the opportunity to work for free on something you own. And to the younger founders reading this, honestly, you have no idea how many good ideas and promising pre-seed startups I have watched quietly shut down over the years. Maybe one percent of you will make it. That is not pessimism, that is the base rate, and everyone in this industry knows it. So here is the deal. It is your idea, your company, your upside, and therefore your risk. If you want professionals to build the thing that makes it real, you carry the cost of building it. That is not me being greedy, that is me refusing to fund your startup with my time while you keep the equity. Equity is a bet we make together. A quote is what it costs to get to the point where there is anything worth betting on. Pay for the work. Then let us talk about the upside. #startups #freelance #founders #buildinginpublic #consulting
-
Vadim Chepegin heeft dit gedeeldI recently looked at ProofLine through a different lens: the UN Sustainable Development Goals. The clearest connection is SDG 8: Decent Work and Economic Growth, especially the objective of productive employment and decent work for all. Finding work is not only about producing a polished CV. People need to understand where their experience fits, which opportunities are worth pursuing, what evidence supports their claims, and where real gaps remain. That is the problem ProofLine is trying to address. It helps professionals organise their real experience, compare it with opportunities, prepare focused applications and keep control over what is written and submitted. It does not guarantee employment, and I do not want to claim an impact we have not measured. But the direction is clear: make the path between experience and suitable work more transparent, deliberate and evidence-based. #UN #ProofLine #SDG8 #DecentWork #FutureOfWork #JobSearch #CareerDevelopment #ResponsibleAI #HumanCenteredAI
-
Vadim Chepegin heeft dit gedeeldThanks to the beta-testers, we are fixing the onboarding flow and simplifying UX every day! #JobSearch #CareerStrategy #ResumeWriting #AIWriting #Hiring #Recruiting #FutureOfWork #TechCareers #HumanFirst #EvidenceBased
-
Vadim Chepegin heeft dit gedeeldI am going to test it and see if it delivers on its promise!Vadim Chepegin heeft dit gedeeldA few weeks ago, I said #Foveance cuts LLM token costs by 60%. I was clearly underselling it! The real number is 82%, lossless, with every benchmark public and reproducible. Here's a cost most AI agent builders quietly absorb: agents re-send their entire conversation history on every turn. The same directory listings, the same stack traces, the same tool outputs, transmitted again and again. "But prompt caching solves that." Partially. Caching discounts repeated tokens, it doesn't remove them. They still consume your context window, cache reads still cost real money, and a single changed character breaks the match. #Compression stacks on top of caching. It doesn't compete with it. Foveance is a true compression codec for agent context, not a summariser. Summarisers discard detail and hope it wasn't important. Foveance works the way gzip does: it detects text that has already appeared and replaces it with a short pointer. Nothing is dropped. unpack(pack(x)) == x, verified by property-based testing. The results, all from real runs and all reproducible on GitHub: By default you get 75% fewer input tokens, and a single flag takes that to 82%. On a real 114,000-token coding session it cut 96% of the tokens in 110 milliseconds. Across 5 models, the compressed context actually answered more accurately than the uncompressed one (0.95 vs 0.90), because removing redundancy helps the model find the signal. I should also mention that it outperforms #LLMLingua-2, the leading #prompt #compressor, which is lossy and drops roughly a third of the facts at comparable compression rates. Integration is one command in front of #ClaudeCode, #Codex, #Cursor, or your own stack: foveance wrap --codec --agentic-codec claude Your #API key still goes directly to the provider. Foveance simply makes the bill smaller. This is part of a broader #opensource programme I've been building around efficient, sovereign-ready AI infrastructure. More of it will be public over the coming months. Try it in your browser, no install and no key required: https://lnkd.in/eGUYYifZ Or just go nuts by: pip install foveance Github repo 👇 #LLM #Compression #Algorithm #Token #AI
-
Vadim Chepegin heeft dit geplaatstHow much time do coaches spend repairing AI-written CVs that sound convincing but contain claims the client cannot properly support? I’m exploring a different workflow with ProofLine. Application content begins with career facts the candidate has explicitly reviewed. Evidence gaps become coaching questions - not fabricated bullet points. Drafting can be assisted, but the coach and candidate retain control over every claim and every submission. That could make tailoring more reviewable and help coaches scale their work without attaching their reputation to generic AI output. #JobSearch #CareerStrategy #ResumeWriting #AIWriting #Hiring #Recruiting #FutureOfWork #TechCareers #HumanFirst #EvidenceBased
-
Vadim Chepegin heeft dit geplaatstQuick Update: the 1st beta-test went reasonably well. Of course we have found a number of issues and ways to improve interaction, UX, and clarity of the flow. We are working on the updated version and looking for the beta testers.
-
Vadim Chepegin vond dit interessantOnly two weeks ago, I posted that there was no M&A-activity this year in the Fleet Management space anymore. Glad to see that something is still happening and that with the right partners, deals can still be closed. Congratulations to everyone involved!Vadim Chepegin vond dit interessantAptean acquires FleetGO®, and valantic led the Technology & AI Vendor Due Diligence. FleetGO has built an integrated end-to-end logistics software suite spanning transport management, warehouse management, fleet management and compliance, supported by proprietary telematics hardware, shared platform services and a unified data layer. What stood out in our work was the breadth of the product portfolio and the way FleetGO connects different logistics workflows, customer environments and data sources into an integrated offering. A particular focus was the platform’s ability to support further growth, both organically and through future integration into a broader software ecosystem. FleetGO’s product architecture, integration capabilities and shared platform components provide a strong foundation for continued expansion and post merger integration. Great work by our team around Christoph Küpper, Anton Herzog and Civan Ögün. Congrats to Eike Andre Eul, Job de Fijter, Floris Jansen, Henry Heyer from MAIN, Ronald van Tiel, Jens Beukers, Jorik Schaillée from FleetGO, and to the Aptean team on the transaction.
-
Vadim Chepegin vond dit interessant#GoPottyDiaries 35: Women need to ask for more money Last week I got to speak at FEM Start’s Capital Booster event. Mic in hand, room full of women building their own thing. My favourite kind of crowd. I talked about the Go Potty journey. How I started, the pivot and the pros and cons of using external capital to build your company. Why? Women often opt for bootstrapping way more than men. Only 1% of VC money goes to women. And that number is actually declining (yes…). Female founders need to ask for money more often. It can help accelerate their growth tremendously. So nice to be on stage together with the inspiring Meron Mekonnen and Emilie Sobels. Public speaking used to scare the sh*t out of me. But once you’ve pitched for a room full of investors, talking to a friendly crowd of women, that’s actually easy… Thank you FEM-START and the wonderful Leonie Ebbes for having me. Great energy, great questions, great event! #GoPottyDiaries #startup #impact #femalefounder #funding ----- 𝘌𝘭𝘪𝘢𝘯𝘯𝘦 𝘓𝘦𝘦𝘧𝘧𝘦𝘳𝘴, 𝘧𝘰𝘶𝘯𝘥𝘦𝘳 𝘰𝘧 𝘎𝘰 𝘗𝘰𝘵𝘵𝘺. 𝘈𝘯 𝘪𝘮𝘱𝘢𝘤𝘵 𝘤𝘰𝘮𝘱𝘢𝘯𝘺 𝘵𝘩𝘢𝘵 𝘨𝘦𝘵𝘴 𝘤𝘩𝘪𝘭𝘥𝘳𝘦𝘯 𝘰𝘶𝘵 𝘰𝘧 𝘥𝘪𝘢𝘱𝘦𝘳𝘴 𝘴𝘰𝘰𝘯𝘦𝘳 𝘵𝘩𝘳𝘰𝘶𝘨𝘩 𝘵𝘩𝘦 𝘦𝘷𝘪𝘥𝘦𝘯𝘤𝘦 𝘣𝘢𝘴𝘦𝘥 𝘎𝘰 𝘗𝘰𝘵𝘵𝘺® 𝘢𝘱𝘱 🌎. 𝘐𝘯 𝘵𝘩𝘦 #𝘎𝘰𝘗𝘰𝘵𝘵𝘺𝘋𝘪𝘢𝘳𝘪𝘦𝘴 𝘐 𝘨𝘪𝘷𝘦 𝘺𝘰𝘶 𝘢 𝘴𝘯𝘦𝘢𝘬-𝘱𝘦𝘢𝘬 𝘪𝘯𝘵𝘰 𝘮𝘺 𝘫𝘰𝘶𝘳𝘯𝘦𝘺, 𝘮𝘺 𝘭𝘦𝘴𝘴𝘰𝘯𝘴 𝘢𝘯𝘥 𝘸𝘩𝘢𝘵 𝘣𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘳𝘦𝘢𝘭𝘭𝘺 𝘧𝘦𝘦𝘭𝘴 𝘭𝘪𝘬𝘦.
-
Vadim Chepegin vond dit interessantVadim Chepegin vond dit interessantWorkshop voor MBO-studenten: AI in de praktijk Gisterochtend mocht ik een workshop geven aan de eerstejaars Handel & Ondernemen MBO-studenten van ROC Amsterdam. Ik vond het echt superleuk om te doen en was erg onder de indruk hoe snel de studenten EU GPT eigen maakten. Na bedrijfsonderzoek via AI websearch en eigen analyse waren er top presentaties op basis van AI-gegenereerde PowerPoint-slides. Deze studenten gaan het verschil maken in bedrijven. Door AI slim in te zetten, kunnen organisaties zich weer richten op hun kern. Zij voegen de waarde toe, zij bouwen de toekomst. Succes allemaal met jullie opleiding! Hayat E.iding! Hayat E., Kevin van de Kamp Bedankt voor deze mogelijkheid!
-
Vadim Chepegin heeft hierop gereageerdVadim Chepegin heeft hierop gereageerdLets briefly talk about M&A-activity in the fleet managment space this year. On 30 June, Vontier closed the sale of Teletrac Navman for a headline 220 million dollars. The quarterly filing tells a more interesting story than the press release. 85 million came as cash. Most of the rest was a seller's note, meaning Vontier lent the buyer a large part of the purchase price. They kept a minority stake in the business they were selling. And they booked a loss of roughly 86 million dollars on the way out. The seller financed a substantial share of its own exit and wrote off 86 million for the privilege. That is the benchmark for M&A activity in the fleet management space in 2026, because it is one of the very few deals that actually landed this year. Everyone blames AI and the SaaSpocalypse. I spent some time working out whether that discount is deserved, and I do not think the answer is simple. Fleet management gets priced as software, but a meaningful share of the revenue is protected by things software multiples do not model: hardware in vehicles, workshop-dependent switching costs, and regulatory liability that somebody has to carry. Some of the discount is just wrong. Some of it was overdue before anyone had heard of an agentic workflow. I have written up three reasons fleet management is different, three reasons the discount is deserved anyway, and the five things I would actually look at in a due diligence process in the AI-age. Link is in the comments. If you are looking at something in this space right now, I would be interested in how you think about what is happening right now in fleet management M&A.
-
Vadim Chepegin vond dit interessantVadim Chepegin vond dit interessantSINTRA surveillance fieldlab has resolved an actual theft case in Port of Moerdijk. While still running as a research ITEA 4 project, the SINTRA system has helped to find the thieves that stole the equipment. Funny enough, the equipment belonging to the project. At that day, the research team of AIMS lab TUe, ViNotion, Sorama, Avular, and MantiSpectra was playing criminal actions for data collection purposes at the private premises of Port of Moerdijk. When the data recordings were completed, we have realized that some of our equipment was stolen. Pity, since it can only be purchased in Japan. But SINTRA was created to target such cases, so the sub-team at Eindhoven University of Technology and Port of Moerdijk has found the thieves in 4 steps. 1️⃣ We have filed a report to the police and checked the recordings from our portable sensors. It became clear that an unknown large van may be involved (see images). It disappeared within similar timeframe with the equipment. 2️⃣ We have reported this to the police and asked our colleagues at the control room of PoM to check this timeframe in the 360 camera installed on the mast nearby. They have confirmed that two unknown persons took the equipment, placed it in the van and departed. 3️⃣ The large van had a company logo, so we reported all details to the police by phone and they promised to call us back. Unfortunately, they never did. So we continued the search ourselves. 4️⃣ Using OSINT technologies, we have found the company owner and his phone number. Several calls confirmed that his workers have stolen our equipment. The owner appeared to be cooperative and in two days the equipment returned home. At the end, I told the owner: Your guys have tried to steal things from a wrong project, since the project is focusing on theft detection. Thanks to the research team members involved in the investigations: Sara Abdulaziz, Michiel van den Berg, Peter VISSER, Dick Scholte and Răzvan Balmoș. P.S. Unfortunately, the police have not shown any signals of collaboration or support. #Security #Surveillance #AI #Sensors #CriticalInfrastructure #Theft #CrimeDetection #Harbour #PortSecurity #TUE #Politie #RVO
-
Vadim Chepegin vond dit interessantNext generation RDF Semantic #KnowledgeGraph Database: I am going to launch the first massively distributed #RDF database with auto-scaling on Kubernetes, to replace #ApacheJena. It is written in #Rust and #C++, who wants to contrib on #GitHub once I launch it? #opensource Barry Smith John Beverley Dr. Lorraine Bassett MBA, PhD Mark Fidelman Pan Wu Alana Podrx Jessica Talisman Ashleigh N. Faith Sumanta M. Malachi Keddington
-
Vadim Chepegin vond dit interessantVadim Chepegin vond dit interessantReady for a new real-world system design challenge? Dive into today’s System Design Interview: Wallet / Ledger! 💸 The hard part of a wallet is not just "subtract here, add there." It is showing, even after retries, crashes, multiple debits at once, provider delays, and data splits, that the system never lost or created money. Try the interactive walkthrough: https://lnkd.in/eCCBDSYk This interview carefully goes through the steps: starting with a simple single-balance column, then switching to an unchangeable double-entry journal, saved retry records, calculated balances, managing simultaneous actions, a transactional message queue, error checking, and clear settlement states for outside systems. The key lesson is that choosing "accuracy over always being available" changes the system design. Fast balance checks are allowed, but the journal is the source of truth. Retries happen, but using the same retry key won't cause duplicate entries. Events can be sent later, but the message record is saved together with the ledger update. Transfers across data splits can use a process, but users must see a clear pending status instead of available money on both sides. Modern technology helps, but it does not change the accounting system. Tools like PostgreSQL, Aurora, Spanner, DynamoDB, Temporal, Kafka, S3/Cloud Storage, KMS, and monitoring tools are all options around the main rule: every entry balances, every calculated view can be recreated, and every outside payment can be checked. Explore the project/book catalog: https://lnkd.in/eApX2Ygj Free source code: https://lnkd.in/e93-t5GU #SystemDesign #SystemDesignInterview #SoftwareArchitecture #FinTech #DistributedSystems
-
Vadim Chepegin vond dit interessantI wrote this post a few weeks ago before I went on holiday. The harsh truth: we went on holiday a week ago. I was ready to switch off mentally and be there for my kids. The reality: I have been working every evening until midnight and trying to work as much as I can during the day. Trying really hard to hide this from my kids. Big fail. They noticed it. Are you on the phone again?? You're on your laptop ALL the time!! They feel I am not mentally present because there are some urgent things to attend to in the company. It breaks my mother heart if they complain. But this summer holiday some things in the company cant wait. Really bad luck. I hope my kids dont hate me for it later 🙈 . #GoPottyDiaries 33: F*ck the algo, it's family time. Always working. Always on. As an entrepreneur, work never really stops. So I caught myself thinking: should I write a few posts ahead so I can keep posting while we're away? Stay consistent, keep the algorithm happy? And then: why am I even thinking about this? F*ck the algo, we’re going on holiday. It's family time. My kids don't care about my posting schedule. They care about whether I'm actually there, phone down, fully present. That's the whole point of this break. Not just rest, but reconnecting. So I come back not just relaxed, but with my head back in the game and full focus on the second half of the year. Just like last year, when I took a complete break. It was so good. And I can do it again, now that I actually have a team around me. If there's a fire, I'll fight it. Until then? Sun, sea and the beach in Sweden. And when we're back, a slow restart. Straight into our very first workcation. The whole team, families included. Not a fancy resort in Ibiza yet, we're a startup, not Google. Just a farmhouse somewhere in Zuid-Holland. But we have to start somewhere, right? For now: see you in a few weeks. ☀️ #GoPotty #GoPottyDiaries #startup #founderlife ----------------- 𝘌𝘭𝘪𝘢𝘯𝘯𝘦 𝘓𝘦𝘦𝘧𝘧𝘦𝘳𝘴, 𝘧𝘰𝘶𝘯𝘥𝘦𝘳 𝘰𝘧 𝘎𝘰 𝘗𝘰𝘵𝘵𝘺. 𝘈𝘯 𝘪𝘮𝘱𝘢𝘤𝘵 𝘤𝘰𝘮𝘱𝘢𝘯𝘺 𝘵𝘩𝘢𝘵 𝘨𝘦𝘵𝘴 𝘤𝘩𝘪𝘭𝘥𝘳𝘦𝘯 𝘰𝘶𝘵 𝘰𝘧 𝘥𝘪𝘢𝘱𝘦𝘳𝘴 𝘴𝘰𝘰𝘯𝘦𝘳 𝘵𝘩𝘳𝘰𝘶𝘨𝘩 𝘵𝘩𝘦 𝘦𝘷𝘪𝘥𝘦𝘯𝘤𝘦 𝘣𝘢𝘴𝘦𝘥 𝘎𝘰 𝘗𝘰𝘵𝘵𝘺® 𝘢𝘱𝘱 🌎. 𝘐𝘯 𝘵𝘩𝘦 #𝘎𝘰𝘗𝘰𝘵𝘵𝘺𝘋𝘪𝘢𝘳𝘪𝘦𝘴 𝘐 𝘨𝘪𝘷𝘦 𝘺𝘰𝘶 𝘢 𝘴𝘯𝘦𝘢𝘬-𝘱𝘦𝘢𝘬 𝘪𝘯𝘵𝘰 𝘮𝘺 𝘫𝘰𝘶𝘳𝘯𝘦𝘺, 𝘮𝘺 𝘭𝘦𝘴𝘴𝘰𝘯𝘴 𝘢𝘯𝘥 𝘸𝘩𝘢𝘵 𝘣𝘶𝘪𝘭𝘥𝘪𝘯𝘨 𝘳𝘦𝘢𝘭𝘭𝘺 𝘧𝘦𝘦𝘭𝘴 𝘭𝘪𝘬𝘦.
Ervaring en opleidingen
Licenties en certificaten
-
Machine Learning With Big Data
Coursera Course Certificates
-
Big Data Integration and Processing
Coursera Course Certificates
-
Graph Analytics for Big Data
Coursera Course Certificates
-
Big Data Modeling and Management Systems
Coursera Course Certificates
-
Introduction to Big Data
Coursera Course Certificates
Ervaring als vrijwilliger
-
Peer reviewer
Educational Technology & Society Journal
- 14 jaar 9 maanden
Opleiding
SSCI indexed, the Journal of Educational Technology & Society has 5-Year impact factor 1,376 according to Thomson Scientific 2014 Journal Citations Report.
Publicaties
-
Turning Your Legacy Systems into Future Profit: Innovation in Production Management
Cutter IT Journal 28(4):21
Production data is usually distributed across various heterogeneous data sources such as legacy systems, event producers, sensors, planning and scheduling tools, etc. Thus, in order to pave the road towards sensing enterprise, there is a need for a robust communication layer, which allows seamless data exchange and integration. As discussed in this article, the proposed solution involves the use of ontologies, well-defined models that describe the important concepts, relationships, and…
Production data is usually distributed across various heterogeneous data sources such as legacy systems, event producers, sensors, planning and scheduling tools, etc. Thus, in order to pave the road towards sensing enterprise, there is a need for a robust communication layer, which allows seamless data exchange and integration. As discussed in this article, the proposed solution involves the use of ontologies, well-defined models that describe the important concepts, relationships, and constraints of the domain of production management. Due to the sensitivity of the aforementioned data, the adoption of a reliable solution for controlling data access and secure communications is essential. An extra challenge of this project lies in the fact that any MAS needs a massive information flow up-front for initial set up and then continuously during its execution to make its work effective. This is a well-known so-called “knowledge bottleneck” problem, but, in the case of ARUM, velocity and volume of data are also playing an important role since a schedule for workers that is based on e.g. the outdated number of screws, connectors, etc. would have no value. This article is focused on the enterprise service and data integration aspects of the proposed platform.
Andere auteurs -
Advances in data value chain management: semantic web and a mixture of integration paradigms for holonic systems
Complex Systems: Control and Modeling Problems
The ARUM platform makes use of Multi-Agent Systems (MAS) that build schedules in real time based on the information collected from a wide variety of resources including legacy systems, data generated by workers, sensors, MAS themselves, etc. to bring the necessary information enterprise integration platforms that supports different paradigms in order to guarantee fast delivery of business value, maximal coverage of systems accepted by the platform, and security. Innovative solution described in…
The ARUM platform makes use of Multi-Agent Systems (MAS) that build schedules in real time based on the information collected from a wide variety of resources including legacy systems, data generated by workers, sensors, MAS themselves, etc. to bring the necessary information enterprise integration platforms that supports different paradigms in order to guarantee fast delivery of business value, maximal coverage of systems accepted by the platform, and security. Innovative solution described in this paper is based on TIE Smart Bridge (TSB), - a lightweight cloud-based integration bus, and TIE Smart Integrator (TSI), - a schema mapping and data transformation tool that intensively use Semantic Web technologies in order to overcome limitations of syntactic based approaches. The solution is delivered as SaaS and business functions are packaged as a set of microservices to speed up a time-to market. This paper focuses at one part of the ARUM platform, namely data transformation and integration aspect of it.
Andere auteurs -
Discovery Service for User Models in a Multi-Application Context
ICALT 2005: 956-957
Andere auteurs -
Ontology-driven User Modeling for Modular User Adaptive Systems
LWA 2004: 17-19
Andere auteurs -
Semantic Web-based Adaptive Hypermedia
WWW Workshop on Application Design, Development and Implementation Issues in the Semantic Web 2004
Andere auteurs -
The Next Big Thing: Adaptive Web-Based Systems
J. Digit. Inf. 5(1) (2004)
Andere auteurs
Cursussen
-
BackBase Client Training Program based on Backbase Client Edition 4 .0 RC, Backbase training center, the Netherlands, 4-8 June, 2007
-
-
CIRCLE Summer School on Intelligent Tutoring Systems, Carnegie Mellon University, USA –2003
-
-
Data Mining – IKAT, Universiteit Maastricht –2003SSSW-2003 The First European Summer School on Ontological Engineering and the Semantic Web, Spain –2003
-
-
Google Cloud Platform Fundamentals
CP100A
-
Introduction Course in the Wholesale Banking Culture, provided by ING Wholesale Banking Academy, 3 – 5 December 2007, Ijmuiden, the Netherlands
-
-
The 1st European Summer School on Ontological Engineering and the Semantic Web (SSSW 2003)
-
-
The Summer School on Service and Software Architectures, Infrastructures and Engineering, June 28 – July 3 2010, Crete, Greece
-
Projecten
-
ACCEPT
-
Key technologies: Spring boot, Java 7/8, RESTful, microservices, Swagger, Docker, NoSQL (Neo4j - core model; Redis - cache and queue, pub/sub; Neo4j - reporting), Twilio (telco API).
Our focus in this project is related to data integration and distribution. The so-called Information Lake should be built on top of isolated data silos in order to
- Provide guidelines and methodologies during the construction process by transferring knowledge between the different stakeholders of the…Key technologies: Spring boot, Java 7/8, RESTful, microservices, Swagger, Docker, NoSQL (Neo4j - core model; Redis - cache and queue, pub/sub; Neo4j - reporting), Twilio (telco API).
Our focus in this project is related to data integration and distribution. The so-called Information Lake should be built on top of isolated data silos in order to
- Provide guidelines and methodologies during the construction process by transferring knowledge between the different stakeholders of the construction process
- Increase the efficiency, reliability, and productivity of construction processes by providing workflows for the interaction between different entities on a construction site
- Improve the final thermal, acoustic and energy performance of buildings by providing sophisticated quality assurance tools, which will be used actively or passively by different stakeholders during the construction process and provide self-inspection capabilities.Andere bijdragersProject weergeven -
SAM
-
Vadim was a deputy coordinator and a technical manager of this project.
SAM provides open, standardised ways of characterising, discovering and syndicating media content items interactively. Users are able to consume and prosume media content from different syndicated sources, using different synchronised devices, ranging from tablets, phablets and smartphones to notebooks and connected TV sets. With SAM content providers can create richer, enhanced experiences around original media content…Vadim was a deputy coordinator and a technical manager of this project.
SAM provides open, standardised ways of characterising, discovering and syndicating media content items interactively. Users are able to consume and prosume media content from different syndicated sources, using different synchronised devices, ranging from tablets, phablets and smartphones to notebooks and connected TV sets. With SAM content providers can create richer, enhanced experiences around original media content items such as television programming, TV shows and on-demand video.Andere bijdragersProject weergeven -
ARUM
-
A large project (€ 12M costs) with the main goal of the optimisation a ramp up for complex small lot production such as Airbus. I was responsible for the system integration work package. We have developed a micro-service infrastructure using a mix of proprietary and open source software. The main technical challenge was in transformation and integration data coming from various sources into a complex ontology. Horizontal scalability was achieved via distribution of the service tier in the cloud…
A large project (€ 12M costs) with the main goal of the optimisation a ramp up for complex small lot production such as Airbus. I was responsible for the system integration work package. We have developed a micro-service infrastructure using a mix of proprietary and open source software. The main technical challenge was in transformation and integration data coming from various sources into a complex ontology. Horizontal scalability was achieved via distribution of the service tier in the cloud with help of automated provisioning and deployment toolset.
Main technologies: Java, SOA, OWL, OpenStack, Cloud, Vagrant, Puppet, Hadoop, TIE Smart Bridge (TSB), TIE Smart Integrator (TSI)Andere bijdragersProject weergeven -
INTUITEL
-
Vadim was involved in the proposal writing, negotiation, and setting up the project phases.
INTUITEL is a research project co-financed by the European Commission with the aim to advance state-of-the-art e-learning systems via an addition of guidance and feedback for learners. Through a combination of pedagogical knowledge, measured learning progress and a broad range of environmental and background data, INTUITEL systems will provide guidance towards an optimal learning pathway. This allows…Vadim was involved in the proposal writing, negotiation, and setting up the project phases.
INTUITEL is a research project co-financed by the European Commission with the aim to advance state-of-the-art e-learning systems via an addition of guidance and feedback for learners. Through a combination of pedagogical knowledge, measured learning progress and a broad range of environmental and background data, INTUITEL systems will provide guidance towards an optimal learning pathway. This allows INTUITEL-enabled learning management systems to offer learners automated, personalized learning support so far only provided by human tutors.Andere bijdragersProject weergeven -
CREATE
-
Creating evolution capable cooperating applications in industrial automation. CREATE developed innovative solutions supporting the seamless integration of manual processes into efficient industrial systems through self-configuring, flexible and evolutionary techniques.
TIE was responsible for delivery an integration architecture (SOA, micro-services) and data integration and transformation. TIE has also delivered a widget-based user-friendly portal that could be used by the end users for…Creating evolution capable cooperating applications in industrial automation. CREATE developed innovative solutions supporting the seamless integration of manual processes into efficient industrial systems through self-configuring, flexible and evolutionary techniques.
TIE was responsible for delivery an integration architecture (SOA, micro-services) and data integration and transformation. TIE has also delivered a widget-based user-friendly portal that could be used by the end users for configuring and monitoring their own workflows.Andere bijdragersProject weergeven -
PREMANUS
-
The goal of PREMANUS is to overcome the asymmetric distribution of information in the End-of-Life (EoL) recovery of products by connecting OEMs and subcontractors, with a special emphasis on remanufacturing.
Andere bijdragersProject weergeven -
OPDM
-
TIE was in charge of the data transformation and integration. Based on the results of the previous successful project called STASIS, TIE developed further TIE Smart Integrator (TSI) and used it for e-Commerce domain in order to fuse data about products received in various formats into consistent semantically enriched view for further processing by personalisation engines.
Andere bijdragersProject weergeven -
OMELETTE
-
Overall Objective: Enabling lightweight composition with focus on communication and collaboration
Purpose: Minimizing effort in creating Web applications, both for end users and developers
Key Ideas:
- User friendly building blocks: UI widgets
- End users compose widgets: workspaces
- Development: systematic Web engineeringAndere bijdragersProject weergeven -
SOA4ALL
-
Vadim was in lead for the demonstrator mini-project within SOA4ALL that built an iPhone/iPad application that was using Open and Linked Data through SOA4ALL semantic web services platform and employed RESTful interfaces for data and services integration into one user-friendly experience.
Andere bijdragersProject weergeven -
NEXOF-RA
-
The overall ambition of NEXOF-RA is to deliver a Reference Architecture for the NESSI Open Service Framework (ranging from the infrastructure up to the interfaces with the end users) leveraging research in the area of service-based systems to consolidate and trigger innovation in service-oriented economies.
In particular the scope of NEXOF-RA is to deliver:
1. the NEXOF Reference Architecture, following an Open Architecture Specification Process which will allow contributions from…The overall ambition of NEXOF-RA is to deliver a Reference Architecture for the NESSI Open Service Framework (ranging from the infrastructure up to the interfaces with the end users) leveraging research in the area of service-based systems to consolidate and trigger innovation in service-oriented economies.
In particular the scope of NEXOF-RA is to deliver:
1. the NEXOF Reference Architecture, following an Open Architecture Specification Process which will allow contributions from many sources also outside NEXOF-RA;
2. a proof-of-concept; this will be a set of software artefacts the project team will use to validate the key architectural choices made;
3. the NEXOF Roadmap; this will define the roadmap for the implementation and adoption of the whole of NEXOF-RA. This will consist of a Strategic Plan, NEXOF requirements, supporting tools such as a wiki-based collaborative working environment, and the NEXOF Repository. The NEXOF Roadmap could be an instrument of exploitation for all the NESSI Strategic Projects and other possible contributing initiatives.Andere bijdragersProject weergeven -
STASIS
-
Vadim was leading the technical integration and development of the overall system. He was also contracted to start up the Java expertise team within the company. Vadim introduced agile methodologies to the internal team (5 members) as well as he scaled that up to the whole dispersed project team (up to 40 people around the Globe including China). Vadim has also set up the practices for continuous integration and test-driven development.
Andere bijdragersProject weergeven
Onderscheidingen
-
PhD student scholarship of the President of the Russian Federation
the President of the Russian Federation
Got it 3 times as a student and 1 time as a PhD student within the period of 1997 - 2001
Talen
-
English
Volledige professionele vaardigheid
-
Russian
Moedertaal of tweetalig
-
Dutch
Basisvaardigheid
Organisaties
-
Big Data Value Association
SME representative in the technical meetings
-http://www.bdva.eu Participation in the technical meetings during the preparation of the original cPPP and SRIA. Preparation of a sectorial industrial workshop (media) with the EC and Fraunhofer Institute.
Ontvangen aanbevelingen
11 personen hebben Vadim aanbevolen
Word nu lid om dit te bekijken