About
Patrick (PJ) Jean is Founder and CTO at Cerebe, where he leads engineering and research.…
Activity
9K followers
-
Patrick Jean shared thisSoftware development is a verifiable domain. That statement is why AI code generation can work at all. If you couldn't check the output against a standard, faster writing would just be faster garbage. Most teams already feel the gap. The model produces. Definition of done still has to be enforced. Without an independent check bound to the change, "approved" is a vibe. Cerebe is the quality engine for that; it works with the coding harnesses you already run; Claude Code, Grok Build, Codex, OpenCode, and the rest; via CLI/MCP and an adversarial, cross-family model critic panel, so every AI generated commit and PR is judged against your standards before it ships.Patrick Jean shared thisMost teams adopting AI coding feel the same tension: go faster, or stay sure about what shipped. Cerebe is built for that gap. We’re the quality engine for AI software development — so companies can move at AI speed without losing confidence in the result. Teams can use our platform in their own workflow, or work with our Forward Deployed Engineering team to build on it alongside them. Learn more at cerebe.ai
-
Patrick Jean shared thisLooking for a deeply technical, business-minded, AI-immersed Chief Architect. This person will own and drive the technology and architectural roadmap for the Employ product portfolio and be a key member of my senior leadership team. If you’re interested in architecting and building the technology behind the products that have helped match millions of people with meaningful work, contact me.
-
Patrick Jean shared thisI'm looking for a deeply technical VP of AI to lead our AI product design, architecture and development. You'll have the opportunity to drive the vision and implementation of AI throughout the product portfolio. If you move fast and you are at the cutting edge of AI reach out to me directly. https://lnkd.in/grz6f8St
-
Patrick Jean posted thisWhen I launched my first startup in college, I was at the start of a lifelong obsession with technology and creating. Over time, that evolved into a deep interest in building in markets where it has the potential to unlock meaningful, untapped impact. With the rise of AI, the hiring and recruiting space is having one of those moments. Teams know the benefits of AI, but they’re still navigating where and how to use it in the best way. That’s why I’m excited to share the next step in my journey: joining Employ as CTO. With experience building teams from the ground up and scaling them during periods of growth, I’ve seen firsthand the complexities and friction that can exist within hiring processes. AI has a real opportunity to help alleviate some of those challenges, for both companies and jobseekers, making the experience and outcome much more meaningful and rewarding. At the same time, recruiting comes with responsibility. Because, at its core, it’s about connecting people with opportunity. That means innovation needs to make hiring smarter and faster—while keeping human connection at the center. Employ already has an incredible foundation for how they strike this balance through a responsible, people-first product philosophy and a commitment to applying AI where it can handle busy work without replacing human judgement. I’m excited to build on that foundation, partnering with the team to strengthen the platform, accelerate innovation, and help shape the future of intelligent hiring. Here’s to building what’s next! #PeopleFirstRecruiting #FutureOfWork #ResponsibleAI #LifeAtEmploy
-
Patrick Jean shared thisWhy agents don't yet seem like small slices of humanity The current wave of AI agents in the industry are 𝙗𝙪𝙞𝙡𝙩 𝙡𝙞𝙠𝙚 𝙩𝙤𝙤𝙡𝙨, 𝙣𝙤𝙩 𝙤𝙧𝙜𝙖𝙣𝙞𝙨𝙢𝙨. To significantly improve, we need to think more like biologists - hippocampus, instinct, survival - and less like computer scientists - RAM and programs. What has become apparent is that current AI architecture is in early days, especially as related to memory. We're in a context window trap. Increasing context windows to million+ tokens doesn't give agents effective memory, it just gives a larger short term buffer. This is RAM vs wisdom. We're in an amnesia loop. Agents are born and die in a single session. The make the same mistake today that they made yesterday, their brain resets every time. RAG is not memory, it's just library search. True biological memory is developmental at it's core. We need a hippocampus, not just a database. A system that explicitly separates episodic memory - what happened - from semantic memory - what we learned. The solution: AI architect and agentic systems that move from "inject context" to "retrieve experience." From store info to harvest memory. An agent shouldn't just read a doc it should remember "the last time I used this library this way I failed."
-
Patrick Jean shared thisWhy agents don't yet seem like small slices of humanity The current wave of AI agents in the industry are 𝙗𝙪𝙞𝙡𝙩 𝙡𝙞𝙠𝙚 𝙩𝙤𝙤𝙡𝙨, 𝙣𝙤𝙩 𝙤𝙧𝙜𝙖𝙣𝙞𝙨𝙢𝙨. To significantly improve, we need to think more like biologists - hippocampus, instinct, survival - and less like computer scientists - RAM and programs. What has become apparent is that current AI architecture is in early days, especially as related to memory. We're in a context window trap. Increasing context windows to million+ tokens doesn't give agents effective memory, it just gives a larger short term buffer. This is RAM vs wisdom. We're in an amnesia loop. Agents are born and die in a single session. The make the same mistake today that they made yesterday, their brain resets every time. RAG is not memory, it's just library search. True biological memory is developmental at it's core. We need a hippocampus, not just a database. A system that explicitly separates episodic memory - what happened - from semantic memory - what we learned. The solution: AI architect and agentic systems that move from "inject context" to "retrieve experience." From store info to harvest memory. An agent shouldn't just read a doc it should remember "the last time I used this library this way I failed."
-
Patrick Jean shared thisLaney Cannon is that rare combination of amazing thinker and doer. She’s bright, articulate, passionate and driven. You’d be lucky to have her on your team part time or full time or anytime.Patrick Jean shared thisI don’t usually talk about my personal life on LinkedIn. This year made that line harder to keep. In the spring, my grandfather was officially diagnosed with dementia and a seizure disorder. It fractured my focus and, honestly, made me feel like the weakest version of my professional self. By summer, things had stabilized and momentum was returning. In early November, I left my role as Chief of Staff at SoSafe , saying goodbye to some incredible colleagues and meaningful work in security. About a week later, we learned my grandmother had an aggressive, terminal cancer. What we thought would be a journey measured in months or years became a matter of days. I had just accepted a Director of Product Operations role at Aiven with Cassio Sampaio — a role & leader I was so excited about, with a clear growth path, based in Lisbon. Ultimately, I had to rescind the offer and leave Lisbon on just a few weeks’ notice to be present for my family. Both of my parents had passed away by the time I was 19, so stepping in for my grandparents wasn’t really a question — they stepped in for me first. I’m deeply grateful, and it’s a privilege to be able to help. That said, after two months (the longest stretch I’ve been unemployed since I was 14), I’m very ready to use my brain the way I know how. As I navigate what caregiving looks like for my grandfather and work toward getting back to Lisbon, I’m open to project-based, consultancy work, part-time, or contract roles — especially those that could grow into something longer-term. If nothing else, send me a pet project you need a fresh set of eyes on. The kinds of problems I’m best at solving — often sitting between product, tech, and leadership — include: * Chief of Staff to CTO / CPO * Strategic planning & execution * Product and engineering operating rhythms * Cross-functional alignment (product, engineering, GTM, exec) * Compliance & security audits & programs * OKRs, metrics, and decision frameworks * Narrative-building for leadership and boards * Turning ambiguity into clear next steps If you or someone you know could use an extra set of experienced hands in this capacity, I’d love to talk. And if nothing else, I’d appreciate this being shared with the right corners of your network. Thank you for reading — and thank you to those who’ve already checked in. You really know you’ve picked great places to work when former colleagues show up during life’s tough moments & from every previous company they have.❤️
-
Patrick Jean shared thisWe’re building the AI memory layer the world’s agents will run on — and I need 1–3 founding engineers to help ship it now. This is not for everyone. High-intensity, AI-native, first-principles, systems-thinking engineers as owners. At momentiq.ai we’re building Cerebe — an AI-native advanced memory + meta-learning platform (cognitive services for agents/products). Attitude (non-negotiable): • Fully embrace AI in how you build and what you build. • Hungry to learn, crazy sense of urgency, ship daily. • Love hard problems and meaningful work; extreme ownership. Aptitude: • Systems thinking — you see end-to-end: customer → product → data → infra. • Full-stack comfort (AI + backend + APIs + UI). • Smart, pragmatic, relentlessly first-principles. First 90 days: • Ship core AI memory primitives (episodic/semantic stores, retrieval policies, compaction). • Build meta-learning loops (evals, feedback, self-improve). • Stand up AI-native SDLC: LLM+agents for tests/review/docs; trunk-based; fast CI/CD. • Instrument cost/latency/quality; hold the bar. Our stack: Python/TypeScript, Postgres, vector stores (Milvus/pgvector), LLM APIs & vLLM, Kubernetes (KIND local, GKE prod), GitHub Actions, Kafka/Redpanda, OpenTelemetry. If this excites you, apply via the link in the first comment or DM me “Cerebe” with your GitHub and 2–3 shipped things you’re proud of. Years of experience not required — mindset and capability are. If you know someone who lives and breathes AI, tag them. I’m looking for people who will help make the impossible happen fast. #hiring #startup #ai #foundingengineer
-
Patrick Jean shared thisClaude's new weekly rate caps will drop on Aug 28. Some Claude Code users should take notice. What's changing: new 7 day buckets for total usage and more importantly a cap on Opus 4 usage. What happened: Anthropic is a victim of their own success, Claude Code usage has skyrocketed. People found a way to exploit the $200 per month Max plan with 24/7 jobs and parallel agents. Impact? If you're using this as intended, probably no real impact. The current 5 hour rolling window still can be limiting on Opus 4 for intense coding sessions. No change to this. Anthropic missed an opportunity here to easily up-charge for value. The usage patterns should be straight forward to detect the less than 5% of users that are doing around the clock or parallel processing, simply make it easy to pay for it instead of API level configuration. Currently Claude Code is my go to pair and agent programmer. This doesn't change anything for me. I use it every day and it generates more code than any other model for me. I suspect most people won't see an impact but Anthropic should provide a simpler solution for this scale out agent need.
-
Patrick Jean liked thisPatrick Jean liked thisSam Altman on one of the highest-return years of his career: no job, exploring his interests, and a lot of people who were nice enough to teach him things. "I read many dozens of textbooks. I learned about fields that I had been interested in. I didn't have any idea that they were all gonna come together in the way they did." "I learned a lot about nuclear engineering. AI was starting to work, so I learned a lot about AI. I learned about synthetic biology. I learned about investing." " I met people that were working on all sorts of different things who were nice enough to talk to me." "If I met someone interesting and it seemed good and needed help, I would just help them. And they would teach me stuff, or they would offer me the chance to invest in their startup later." "Out of all of it, almost all of it didn't work out. But the seeds were planted for things that worked in deep ways later." With Craig Cannon (Nov 2018)
-
Patrick Jean reacted on thisPatrick Jean reacted on thisKind of chic to represent Corelight for the first time at a customer success leadership conference.
-
Patrick Jean liked thisPatrick Jean liked thisIf you missed last week's Grok Bot Galaxy Livestream, Grok has shared the entire live stream as well as specific sessions. https://lnkd.in/eKZc6v-h
-
Patrick Jean liked thisPatrick Jean liked thisI’m very excited for everyone to try Claude Opus 5.5. It performs at the level of Fable 5.1 on most work, costs 40% less to run than Opus 5, and it achieved the strongest scores of any of our models on our most comprehensive test of model alignment, behavior, and safety. Before its release, Opus 5.5 also underwent pre-release testing by external evaluators including Frontier Design and METR. Sonnet 5.5 and Haiku 5.5 (coming soon!) will feature many of the same improvements to performance, efficiency, and safety. At Anthropic, we believe fulfilling our mission means keeping model capability and safety advancing together, and Opus 5.5 is another step forward. For more information, including how this release fits with our call to pace the frontier, check out the post below. https://lnkd.in/eUK8HwnF
-
Patrick Jean liked thisPatrick Jean liked thisI’m excited to share that I’ve joined Snyk as SVP, Global Partnerships. In many ways, this feels like a natural convergence of the different parts of my career. I spent much of it building businesses and partner ecosystems across enterprise technology, cloud, and software. Over the past five years, I moved much closer to the application of AI — working with industrial AI companies, deploying AI into real production environments, and more recently helping companies think through the technology, economics, partners, and operating models required to turn AI ambition into reality. That experience has reinforced something I believe strongly: AI isn’t simply changing what we build. It is fundamentally changing how we build it. That is a big part of what makes the opportunity at Snyk so exciting. As AI and autonomous agents dramatically accelerate software development, the way we think about security has to evolve with them. Snyk has an extraordinary foundation in developer security, combined with an ambitious vision for securing this new AI-native world through the Snyk AI Security Platform and Evo. I’m equally excited about what this means for our partners. My job is to help make Snyk’s global partner ecosystem an even greater engine of customer value and growth. The opportunity isn’t simply to do more business through partners — it is to grow with them: creating new solutions, services, routes to market, and customer outcomes together. I’ll be reaching out to many people across my network in the coming weeks. I’m also hiring, and I’m looking forward to engaging with existing and new Snyk partners around the world as we build the next stage of growth together. If you work with Snyk today, want to work with us, or see an opportunity to build something together, I’d love to hear from you. Excited to get started! #Snyk #GlobalPartnerships #PartnerEcosystem #AI #Cybersecurity #Growth
-
Patrick Jean reacted on thisPatrick Jean reacted on thisI’m incredibly excited to share that I’ve officially started at Uber! I’m leading Tech and Product hiring, along with uBecome, Uber’s global university hiring program. The scale, pace, and ambition of Uber are inspiring, and I’m energized by the opportunity to help shape the talent and teams that will drive our future. After 25 years in recruiting and incredible chapters at Microsoft and Amazon, I wanted my next chapter to challenge me, keep me learning, and give me the opportunity to make a meaningful impact. From my earliest conversations, Uber stood out because of the people, the ambition, and the belief that exceptional talent is critical to our success. I’m grateful for the warm welcome and couldn’t be more excited to get to work! 🚗✨ If you’re looking for an extraordinary opportunity, take a look at Uber Careers: https://jobs.uber.com #WhatMovesUs #TalentAcquisition #Uber
-
Patrick Jean liked this"Okay, I have a lot to say!" Just wow! You could not ask for a more perfect opening line (OK, maybe "One small step for [a] man...") Congratulations to the Paradromics team and Matt Angle for bringing their amazing vision to life, giving voice and communication back to those who have been shut off from normal pathways. And it is just the beginning. Truly the dawn of a new era. #braincomputerinterface #BCIPatrick Jean liked thisWhen we implanted the Connexus® Brain-Computer Interface (BCI) in June, we crossed a threshold into long-term human use. Today, we crossed another: The first participant in our Connect-One study is now choosing her own words and using neural activity to communicate in real time. She has a progressive motor neuron disease that has significantly affected her ability to speak. Through the Connexus BCI, she answered open-ended questions, spoke freely, and called her grandchildren, asking them questions and responding in a voice generated by the system. Asked what she would tell someone considering this research, she said: "Don't be afraid. It's a wonderful experience to be able to help others by your participation in the research study." Every word she intended was decoded exactly as she meant it. To our first participant: Thank you for your trust, your courage, and the words you have chosen to share with all of us. Our focus begins with restoring communication for people who have lost the ability to speak. Today, it meant that one woman could say what she wanted to say, in her own words, to the people she loves. Read the full release at the link in the comments below. CAUTION: The Connexus® BCI is an investigational device limited by United States law to investigational use.
-
Patrick Jean reacted on thisPatrick Jean reacted on thisI recently caught up with my mentor and engineering guru, Ken Olson, co-founder at Pegasystems, now retired. Ken taught me virtually everything worth learning about software engineering: craftsmanship, systems thinking, and a love of solving hard problems. Ken’s (intimidating) brilliance and fiery personality is the stuff of Pega legend! I’m eternally grateful for the time, patience, and belief he invested in me. I wish every young professional starting their career journey a mentor like Ken.
-
Patrick Jean reacted on thisThe Omacom Foundation is now officially on LinkedIn! We'll be mirroring all news announcements there. Good page to follow if you care about the future of computing and the first truly agentic operating system!Patrick Jean reacted on thisOmacom Foundation launches with $18.6 million 🚀 It's time to dream big. Omarchy Quattro has given people a chance to experience what the malleable computer of the future looks like, and they like it (a lot!). Making that future broadly available now feels like a moral obligation. To do just that, David Heinemeier Hansson has incorporating the Omacom Foundation: a nonprofit that will hold the trademarks, fund the infrastructure, promote the work, and support the open-source projects and developers Omarchy depends on. Twelve Founding Patrons are each contributing $1 million to the mission: • Tobias Lütke, CEO of @Shopify • Patrick Collison, CEO of @Stripe • Michael Dell, Chairman and CEO of @Dell Technologies • Jack Dorsey, Block Head and Chairman of @Block • Matthew Prince, CEO of @Cloudflare • Brendan Iribe, cofounder of @Sesame and Oculus • Jason Fried, CEO of @37signals • Drew Houston, cofounder and co-CEO of @Dropbox • 🦄 Peter Steinberger, creator of @OpenClaw • Brian Armstrong, CEO of Coinbase • Yunjie Dai, cofounder of XD/TapTap • David Heinemeier Hansson, creator of Omarchy Founding Corporate Patrons, each contributing $1 million a year for three years or $1.5 million in tokens: • DigitalOcean • Meta Superintelligence Labs Distinguished Patrons, each contributing $100,000: • Ryan Hughes, partner at Oodle • Ed Huang, cofounder and CTO of PingCAP - Japan • Adrien Treccani, founder of Metaco • Max Schoening, Head of Product at @Notion Distinguished Corporate Patrons, each contributing $100,000 a year for three years or $150,000 in tokens: • 1Password • 37signals • Four Technologies • Anthropic • Fireworks AI • OpenAI • OpenRouter Over 700 donors have also contributed roughly $120,000 through open patronage, bringing the total in pledges and donations to approximately $18.6 million. We're going to make the prophecy of The Year of Linux on the Desktop come true. All the pieces are now in place. Time to go all in! -- dhh, originally announcing on August 14, now on LinkedIn! https://omarchy.org/
Experience
Education
Licenses & Certifications
-
Microsoft Certified Professional - Windows Server 2012
Microsoft
Issued -
ITIL v3 Expert
EXIN
Issued -
ITIL v3 Service Offerings and Agreements
EXIN
Issued -
ITIL v3 Planning, Protection and Optimization (PPO)
-
Issued -
ITIL v3 Release, Control and Validation
EXIN
Issued -
ITIL v3 Operational Support and Analysis (OSA)
-
Issued -
Microsoft Certified Solution Developer MCSD
Microsoft
Issued
Courses
-
Amazon AWS Architect
-
-
Configuring, Managing and Maintaining Windows Server 2008
6419
-
Planning, Deploying and Managing System Center Configuration Manager 2007 (SCCM)
6451
-
Troubleshooting and Supporting Windows 7 in the Enterprise
6293
Projects
-
Azure IOT
-
Smart building IoT project collecting environmental data from embedded sensors in various devices, transmitting data to Azure Event Hubs, collection in blob storage, real time analysis using Stream Analytics and dashboarding using PowerBI.
-
Microsoft Private Cloud POC
-
System Center Virtual Machine Manager 2012 SP1 POC implementation with App Controller on Hyper-V 2012. SCVMM and SCAC.
-
FlexPod deployment
-
Deploying FlexPod solutions for rapid private cloud buildout.
-
Oracle VM Private Cloud
-
Private Cloud Oracle VM infrastructure buildout to support multiple Oracle application installations including Hyperion, OBIEE, and EnterpriseOne. Buildout of 5TB+ of memory and 500 vCPU environment running Oracle VM backed by EMC CX4 and VNX SAN.
-
System Center Service Manager 2012
-
SCSM 2012 implementation including SCCM, SCOM and AD integration. Incident Management, Change Management and CMDB. Compliance management through logical access and change management processes in SCSM.
-
Oracle Database Appliance ODA
-
Implementation of Oracle Database Appliance ODA for rapid database provisioning and performance improvements. Speed implementation of mid-tier databases while reducing the overhead of ongoing management of Oracle databases. Continue shifting the focus to innovation while reducing overhead operations.
-
Data Protection Manager DPM 2010
-
Implementing disk-based backup solution for Microsoft products using DPM 2010. Replacing tape-backup solution that took 6+ hours for restores, bring down restoration times to minutes. Leveraging licensing from Microsoft ECI. Implementation includes DPM chaining for branch office to central data center and eventually cloud storage. Backup of Microsoft Windows Servers, Windows clients, Linux Servers. Used DPM chaining for site to site backup.
-
Office 365 Implementation
-
Migration of managed Exchange 2007 to Office 365.
-
Mobile Device Manager
-
Implementation of Symantec Mobile Manager software (formerly Odyssey Athena) for managing iPhones, iPads, Androids, BlackBerries, and Windows Phone devices. Integration with System Center Configuration Manager.
-
Oracle EnterpriseOne Implementation
-
Infrastructure implementation for Oracle EnterpriseOne project. Complete buildout of highly available virtual environment on OracleVM, Oracle RAC and EMC SAN.
-
Disaster Recovery Solution - EMC VNX and RecoverPoint
-
Implementation of EMC RecoverPoint for replication of VM and Oracle DB LUNS between EMC CX4 and VNX for disaster recovery.
-
System Center Service Manager 2010
-
Microsoft System Center Service Manager 2010 implementation. ITSM process improvements for new SCSM implementation including Incident Management, Change Management, Problem Management, CMDB and Service Catalog. Integration with SCOM and SCCM for event management and CMDB automation.
Recommendations received
11 people have recommended Patrick
Join now to viewOther similar profiles
Explore more posts
-
Todd Blaschka
Good coverage of what's actually happening here — not just funding news, but a real explanation of why the semantic layer matters: grounded reasoning, auditability, and a meaningful reduction in token costs. That's the part I keep coming back to. This isn't a bet on hype, it's a bet on what AI agents actually need to work in production, from a company that's been building toward this moment for years. #AI #EnterpriseAI #KnowledgeGraphs #SemanticAI #GraphRAG #Graphwise
-
Holger Mueller
.@MongoDB integrates Voyage 4 models across its platform https://bit.ly/49ndKil MongoDB said it is integrating its Voyage 4 embedding and reranking models into its platform infrastructure in a move that will improve accuracy, deliver better context and help developers scale applications.
-
Jojar Dhinsa
THE HYPERSCALER MODEL WAS BUILT FOR THE CLOUD ERA — IT WAS NOT DESIGNED FOR THE AI SOVEREIGNTY ERA AND THE CRACKS ARE SHOWING The hyperscaler model transformed computing. It delivered scale. Efficiency. Accessibility. Global reach. For the cloud era, it was extraordinarily successful. But AI is creating a different set of requirements. Increasingly organisations are asking questions that were not central to the cloud revolution. Questions such as: * Where is the data stored? * Who controls the infrastructure? * Which laws apply? * How resilient is the supply chain? * How do we maintain sovereignty? These concerns are becoming particularly important in: * Defence * Healthcare * Government * Financial services * Critical infrastructure The challenge is not that hyperscalers are failing. The challenge is that the operating assumptions of the cloud era are changing. AI is making infrastructure strategic. And strategic infrastructure often requires greater control. Greater transparency. Greater sovereignty. This is where TerraVaultis MDCaaS provides an alternative deployment model. One that combines AI capability with sovereign-ready infrastructure. Because the future of AI may require something different from the future of cloud. The next decade may not be defined solely by scale. It may be defined by control. #TerraVaultis #MDCaaS #ModularDataCenters #AI #DataCenters #Infrastructure #PrivateEquity #SovereignAI #LiquidCooling
2 Comments -
Charles Lamanna
Excited to see GPT 5.1 already available for use inside Copilot Studio, alongside OpenAI's release today. These experimental models give you the opportunity to evaluate performance against your use cases and existing models to prepare for deployment. Copilot Studio provides model options - from OpenAI, Anthropic, and open-source in Foundry - and tools to decide on the best model for your agent. Looking forward to seeing what people build! https://lnkd.in/gCPKP8kX
14 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content