Sign in to view Valentina’s full profile
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Valentina’s full profile
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
London, England, United Kingdom
Sign in to view Valentina’s full profile
Join with email
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
527 followers
500+ connections
Sign in to view Valentina’s full profile
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Valentina
Join with email
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Valentina
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Sign in to view Valentina’s full profile
or
Already on LinkedIn? Sign in
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
Activity
527 followers
-
Valentina Manea liked thisValentina Manea liked thisWatching the tech industry unleash unpredictable software into production in the name of AI adoption, then spend twice as much time and money trying to limit the blast radius, proves humans should be investing in our own intelligence before allowing the machines to mimic us.
-
Valentina Manea liked thisValentina Manea liked thisAcum un an, alături de Radu Pascale, am pornit un proiect ambițios: să construim un smartwatch de la zero, cât mai rapid, folosind componente off-the-shelf. Rezultatul ne-a surprins chiar și pe noi. Proiectul a demonstrat cât de mult s-a schimbat ritmul în care se poate inova: lucruri care, acum zece ani, necesitau un start-up de 30 de oameni, pot fi realizate astăzi de o echipă mult mai mică, cu resurse incomparabil mai reduse. Poate tocmai de aceea, făcând legătura cu experiențele de atunci, am ales în glumă să-i spunem Hacktor Watch. Sâmbăta aceasta organizăm primul hackathon Hacktor, în cadrul căruia vom porta Zephyr OS pe acest proiect. Dacă vă interesează, încă vă puteți înscrie aici: https://luma.com/bfpmdne1 Locurile sunt limitate.
-
Valentina Manea reacted on thisValentina Manea reacted on thisOpened our State of Alternative Data event in London this week, followed by a great panel discussion with Mark Pason, Mark Thompson, CFA and Tushara Fernando. Good to connect with clients on what’s working, where we are, and how the space is evolving. Thanks to our fantastic team for putting this together!
-
Valentina Manea reacted on thisValentina Manea reacted on thisLooking forward to discussing alternative data and trading with unstructured datasets tomorrow alongside Andrew Simpson, Jon Jenkins, Peter Hafez, Petros Kyliakoudis and Samuel Livingstone, CFA. If you’re attending, would be great to connect. Event details: https://lnkd.in/eNQb_yug #FIXTradingCommunity #EMEATradingConference
-
Valentina Manea reacted on thisValentina Manea reacted on thisToday I left Bloomberg after almost 10 wonderful years. It was an incredible ride. We transformed Python at Bloomberg from the ground up: built state-of-the-art tooling, shipped the latest interpreters on day one year after year, and created what became Bloomberg's most popular open source projects that now are used by thousands of developers worldwide, every single day. Apart from Python I got to work on all the cool things I always wanted to work: from distributed systems to profilers, debuggers, tools that modify ELF files directly, linkers, DWARF parsers and all that dark, low-level magic most people avoid. The problems where you genuinely don't know if something is possible until you make it work. And of course, making Python the language better, faster, and cooler for everyone! I came for the work but I'm leaving with some of the closest friends I've ever had. Very excited about what's next. More soon ;)
-
Valentina Manea liked thisValentina Manea liked thisAs 2025 draws to a close, the disinformation landscape looks fundamentally different than it did just twelve months ago. We’ve witnessed a significant escalation in hybrid threats across industries, with 40% of Refute-monitored mines around the world under active attack from inauthentic actors, and multiple European elections targeted by coordinated campaigns. Threat actors are evolving their tools, including blending bot farms with influencers and deploying agentic AI for full automation. State-sponsored threats are no longer just targeting governments, but undermining corporations and citizens alike. To thrive in this new world order, organisations must get on the front foot. Refute has built world-leading capabilities to help them do this, and a Europe-wide team to meet the disinformation challenge head-on. In 2025, we uncovered threats against the expat community during the Romanian election, discussed European resilience on Euronews, and provided a military-grade AI solution for organisations to stay ahead. We also secured key government and private sector contracts, hosted industry events for important conversations, and published our first research report. We’re proud of our progress and energised to continue scaling our impact in 2026 and beyond. Wishing you and your loved ones a happy holidays and a restful break before the new year. Tom Garnett and Vlad G., co-founders of Refute
-
Valentina Manea liked thisPutting together this video to share my story was an incredible experience. I feel lucky to be part of a team that encourages curiosity and growth, while building the systems that enable our business to succeed.Valentina Manea liked this"It's a partnership. We're building the tools that enable success." In a culture that values every team member’s role and contribution, Nick's curiosity led him to seek more opportunities, learning and delivering in equal measure. His swift ascent is due to prioritizing outcomes over outputs.
-
Valentina Manea reacted on thisValentina Manea reacted on thisMy talk at Open Source Summit is now up! https://lnkd.in/dxNjK7dyGuiding Newcomers: A Mentor’s Journey Through Linux Kernel Development - Daniel BalutaGuiding Newcomers: A Mentor’s Journey Through Linux Kernel Development - Daniel Baluta
Languages
-
English
-
View Valentina’s full profile
-
See who you know in common
-
Get introduced
-
Contact Valentina directly
Other similar profiles
Explore more posts
-
Sawan Shah
Sorcer • 10K followers
Redis has officially returned to open-source licensing with the release of Redis 8 under AGPLv3, marking a significant shift from the controversial SSPL licence that divided the community for years. The decision represents a victory for open-source advocates, driven by internal discussions at Redis and the recognition that SSPL failed to gain acceptance from both the Open Source Initiative and the broader software community. Why This Matters for Rails Developers: 🔹 Redis remains the most popular caching and session store for Rails applications 🔹 AGPLv3 provides clearer licensing terms than the ambiguous SSPL 🔹 Open source status ensures continued community contributions and ecosystem growth 🔹 Rails developers can confidently build on Redis without licensing concerns 🔹 Vector Sets data type and other Redis 8 features are available under true open source terms Redis creator Antirez played a key role in advocating for the licence change, expressing his commitment to writing open source software and ensuring his new Vector Sets implementation would be truly open. His influence helped build internal momentum for the switch across the Redis company hierarchy. Rails applications heavily rely on Redis for caching, background job processing with Sidekiq, and session management. The return to open source licensing removes the uncertainty that enterprise Rails teams faced when evaluating Redis for new projects. Redis 8 also brings performance improvements and new features alongside the licensing change, reinforcing Redis's position as the go-to data structure server for modern web applications. What role does Redis play in your Rails application architecture? We're connecting with Rails developers who understand distributed caching and performance optimisation strategies. #Redis #OpenSource #RubyOnRails #CachingStrategy #TechRecruitment
5
-
KubeFM
7K followers
How do you convince an organization to let automation manage Kubernetes workloads at scale? Yasmin from CloudBolt explains why the real challenge isn't the tooling — it's building enough trust to act on it. Data points, guardrails, preview modes, rollout gates. "You can only go as fast as the speed of trust." Watch the full interview: https://ku.bz/bvr2-nv4P Meet the CloudBolt Software team at Booth 438 at KubeCon Europe 2026 in Amsterdam 👉 https://ku.bz/5pwF68l6Q
36
-
Linearloop
12K followers
Your CI/CD bill is not exploding because of “expensive runners,” it is bleeding out through flaky tests, over-parallelized jobs, and slow feedback loops that quietly tax every deploy This piece shows you how to redesign pipelines so you cut compute waste, shrink queues, and speed up developer flow without adding yet another approval gate or YAML religion Full breakdown in the comments. #DevOps #CICD #PlatformEngineering
8
1 Comment -
Mayur Kumbhar
Addepar • 3K followers
Excited to share blog I wrote at Addepar on using jOOQ for more precise, SQL-first database interactions. At scale, moving away from heavy ORM abstractions helped us gain better control, predictability, and performance. Would love to hear how others are approaching this trade-off. #SoftwareEngineering #BackendDevelopment #Java #jOOQ #DatabaseDesign #SystemDesign #TechEngineering
36
1 Comment -
Aaron Trevena
Amtivo Group • 629 followers
JSON-structure is a modern practical alternative or replacement for JSON-Schema by Clemens Vasters - it's a draft IETF standard and the SDK now includes a Perl implementation after some not so gentle nudging. Clemens has been enthusiastically responsive to feedback and knows his stuff, if you're working with JSON and Perl (or other languages) this is well worth taking for a spin and sending any bug reports or suggestions.. code is at https://lnkd.in/e5fZvw_F and I'm looking forward to seeing the perl library on cpan shortly. #JSON #perl #javascript
8
-
Andy Kuszyk
Typeform • 3K followers
I've been doing some research lately into the Model Context Protocol (MCP), and how to build MCP enabled products at scale. As part of my research into the protocol, I wanted to get a good understanding of how it operated at the network level. `tcpdump` was really handy for this, and I've written up a blog post about how I used it to inspect MCP traffic at the network level to understand the "nuts and bolts" of the client/server communication. If you're interested in doing the same, you can find the write-up here: https://lnkd.in/eVyakGpM #mcp #tcp #ai #debugging
42
2 Comments -
Vroble.com
25 followers
For advanced developers pushing the boundaries of system performance, **Processor Affinity** is a critical, often overlooked, optimization technique. By strategically binding processes or threads to designated CPU cores, engineers can significantly enhance cache utilization and minimize the performance overhead associated with frequent context switching. In high-throughput and latency-sensitive environments, such as real-time trading platforms or scientific simulations, implementing processor affinity can yield substantial gains, with some benchmarks showing up to a 29% improvement in throughput. This level of fine-grained control over CPU scheduling is instrumental in achieving predictable and peak performance. Consider integrating processor affinity into your performance tuning strategy for truly demanding workloads. It's a foundational principle for unlocking your hardware's full potential. #PerformanceEngineering #HPC #SystemsProgramming #DevOps #LowLatency #Optimization
-
Matt Hobbs
1K followers
At the weekend, I published a new write-up on squeezing even more performance out of static sites built with Eleventy and deployed on Cloudflare ✨ In this latest post, I take the idea of compression further by pre-compressing HTML to the highest Brotli level (11) at build time and serving it at the edge via a Cloudflare Worker instead of relying on on-the-fly compression. Why it matters: • Even small edge gains matter at scale – smaller HTML means faster Time to First Byte and quicker overall loads. • Pre-compressing during the Eleventy build shifts compute away from the edge and into your deployment pipeline. • A lightweight Cloudflare Pages Function does smart content negotiation so browsers that support Brotli get the most efficient assets. I walk through how this works, the quirks of serving pre-compressed HTML alongside normal files, and how it fits into a modern edge-centric deployment. If you care about taking static site performance that extra mile, this might be worth a look. https://lnkd.in/e5iM62Sh
20
4 Comments -
Philip Wise
Snyk • 732 followers
The UK wanted a digital strategy. We were just gifted one: Offer a visa to everyone in tech who currently holds a H-1B. Decision in a week. All the top companies have a UK presence already, and it would net us tens of thousands of excellent people who have trained at the best companies. "Opportunities come infrequently. When it rains gold, put out the bucket, not the thimble" - Warren Buffet
123
23 Comments -
Raju Alluri
Salesforce • 6K followers
🚀 Visualizing Flows with Sankey Traces Ever wondered how to clearly show how inputs flow into processes and branch out into outputs? That’s where Sankey diagrams shine. 🔹 Sankey traces map flows between nodes with widths proportional to their magnitude. 🔹 matplotlib has very limited support — but Plotly makes it interactive and powerful. 🔹 All you need: define nodes, links, and values, then build the diagram in a few lines of code. I’ve shared a step-by-step guide (with code from my GitHub notebook ) to walk through the process. 📊 Perfect for customer journeys, process flows, or resource transfers. 👉 Check out the full blog post for details! - https://lnkd.in/gGXucm-9 or https://lnkd.in/gBcrUE3h
28
3 Comments -
Adam Warski
SoftwareMill • 9K followers
What kind of guidance does an LLM need to write a direct-style #Scala 3 application? At the baseline - not a lot, e.g. Claude is quite good both in Scala 3 & direct-style. For finer details - some additions to the prompt might be useful. Read more: https://lnkd.in/dxwy-WdD
20
5 Comments -
Gatikrushna Sahu
Atlassian • 1K followers
Most in-depth technical walkthrough of KIP-1150 What Diskless Topics (KIP-1150) - Diskless Topics = no local broker disk for user data. Instead, data is written directly to cloud object storage (e.g. S3, GCS). - Cost savings are dramatic — up to 80% reduction in TCO (storage + cross-zone network costs) for many cloud-deployed Kafka clusters. - Instant elasticity & autoscaling: since data isn’t pinned on brokers, you can scale brokers up or down quickly without rebalancing data segments or copying logs. #ComputeStorageSeparation #DecoupledArchitecture #CloudNativeData #StreamingArchitecture #ScalableDesign #ModernDataInfra #DataEngineering #DistributedSystems
2
-
Ben Clark
I'm currently researching and… • 1K followers
Exciting to see the Arm ecosystem continue to expand with news of their latest partnership with Octopus Deploy. Together Arm and Octopus are working together to transform how developers build, test, and deliver software. By bringing Arm-based infrastructure into Octopus’ unified delivery platform, teams can manage deployments across architectures seamlessly—improving reliability, accelerating time to market, and enabling true multi-architecture agility. It’s an important collaboration that empowers developers to focus on innovation rather than infrastructure! 💪 https://okt.to/wESlo2
4
-
Baidyanath Prasad
Glance • 6K followers
CAP trade‑offs don’t start in theory class; they start the moment your data leaves a single SSD and has to be agreed on across a network. In Part 2 of “The Physics of Databases”, I break databases into two layers—the local storage engine vs the distributed consensus protocol—and show how the same LSM‑Tree can behave very differently in Cassandra vs Spanner depending on how the network is wired. The post also brings together popular databases (Postgres, Spanner, DynamoDB, Cassandra, Bigtable, Cosmos DB, HBase, BigQuery/Snowflake) into an information‑rich comparison, showing how their storage + consensus choices should shape your primary keys and data‑modelling strategy. 🔗 Read Part 2: The Physics of Databases – The “Two‑Layer” Secret to Navigating the CAP Theorem Link in the comment section👇 💬 I’d love to learn from your side: What stories or hard‑won lessons do you have from choosing between “availability first” vs “consistency first” databases in real systems—and how did those decisions show up later in production or during incidents? #SystemDesign #Databases #DistributedSystems #CAPTheorem #StorageEngines #Cassandra #Spanner #Scalability #BackendEngineering #Architecture
5
1 Comment -
Marco Cuturi
Apple • 3K followers
✨✨✨ We have been working with Michal Klein at Apple MLR on pushing a module to train *flow matching* models using JAX. This is shipped as part of our new release of the OTT-JAX toolbox (https://lnkd.in/etJYGjPv) We do so by also introducing the two changes we advocated over the summer to improve FM training, using either (very) large sharp Sinkhorn couplings (https://lnkd.in/eJ6DnF-U) or, even better and efficient, using our semidiscrete coupling approach (https://lnkd.in/e2aBg27a), proposed with Alireza Mousavi-Hosseini and Stephen Zhang. The tutorial to do so is here: https://lnkd.in/e-BAVejK The tutorial doesn't cover yet the conditional FM problem, but the code does, so give it a try, and as always, we welcome your feedback!
205
2 Comments -
TheNextGenTechInsider.com
584 followers
OpenClaw Resolves Cron Bloat with Efficient Cleanup Strategy 📌 OpenClaw’s performance crisis - caused by cron job sessions bloating the context window - is now fixable with a simple, automated cleanup strategy. Session bloat used to cripple inference speed and memory usage, but new config settings and CLI tools let admins prune stale data instantly. This isn’t just a patch - it’s a game-changer for ops teams running heavy automation pipelines. 🔗 Read more: https://lnkd.in/d4xuDgsx #Openclaw #Croncleanup #Sessionbloat #Inferencespeed #Automatedtasks
-
Dominic Fox
NatWest Commercial and… • 863 followers
This evening's work for Claude, on Patches: 1) Thinking over a couple of API decisions I recorded yesterday, I realised there were cleaner approaches available to one particular aspect, piggy-backing on previous changes. I proposed those approaches to the LLM, talked them through, considered objections, settled on a new approach; the LLM wrote up the revised ADR (which I read, checking that it captured my intention clearly) 2) The method that compares a new patch graph to the existing one and builds an updated execution plan is huge, nested and complex. I would never have written it that way myself, but it works as it stands. Nevertheless, I have future changes targeting that area, and it's risky to have something that's both syntactically convoluted and hard to read and test, especially in such a central piece of logic. It badly needs some attention. I asked the LLM to propose a breakdown of the method into smaller, testable pieces. Its proposal was ok, but missed an opportunity to separate cleanly the "deciding what to do" phase from the "acting on the decision" phase, leaving graph analysis and new module instantiation somewhat tangled up with each other. I suggested some further re-organisation to disentangle things, and asked the LLM to remark on the impact this would have on testing; it confirmed that it would make the testable surface of each of the distinct pieces smaller and more easy to control. Once we had a scheme that looked sound to me, I asked it to write up an epic and a sequence of tickets for it. Then I asked it to start working through the changes. Now it's chugging through the first ticket, while I write this. It will go on doing so while I pop downstairs and do the washing-up. I'm comfortable with the fact that the first draft of the code it's now tidying up was unsatisfactory from a testability and maintainability point of view: it grew organically and expediently in the course of pulling together something that worked well enough for me to start playing with it. Now it's a glaring area of technical risk, so I'm cheaply and quickly addressing it. A capable human developer would have kept much better discipline throughout the drafting process, but might not have seen that the decide/action split was natural and desirable until about this point in the game - I didn't until I stopped to think about it. What's absent from this process, for me, is anxiety that the LLM will just heap up piles of intractable slop. It will typically do the expedient thing at each step along the way, but that's fine if you're working in short iterations provided you're also periodically stopping to review where cruft and friction are accumulating. What it lacks in foresight, you somewhat have to make up for in hindsight. But the latter is famously somewhat clearer than the former anyway.
2
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore More