This month, we’re focusing on what keeps clients with us—reliability, results, and readiness for what’s next.
We’re proud to share a new testimonial from Microsoft Azure on how Akvelon grows projects into long-term partnerships, along with a look at the kinds of challenges we solve for companies like Microsoft, Google, GitHub, and Reddit.
In our latest post series, we share practical takeaways on whether AI agents can replace apps, running real enterprise tasks with no active backend, and minimal prompting - and where the limits still are.
Let’s dive in:
Five Years of Trust With Microsoft Azure
What starts as a single project can grow into a long-term, trusted partnership.
In a new video testimonial - LinkedIn post, Michael Lu, Principal Software Engineering Manager at Microsoft Azure, reflects on our 5+ year collaboration and how it continues to evolve, from front-end development to data pipelines, ML, and AI projects.
Michael emphasizes the proactive engineering, smooth execution, and reliable collaboration that scaled alongside Azure’s needs while keeping projects on track.
We’re honored to support the Microsoft Azure team with technical depth and a partnership mindset, and grateful to Michael, who trusts us to contribute to their long-term success.
Engineering Solutions for High-Stakes Challenges
Microsoft. Google. GitHub. Reddit.
These are just a few of the companies that rely on Akvelon when the stakes are high and the constraints feel impossible: tight deadlines, untouchable legacy systems, and teams that are already stretched thin.
On our new services webpage, we’ve outlined how we’ve supported these and other tech leaders, helping them ship, scale, and evolve the platforms their businesses run on.
What we build together:
– Systems that stay reliable as they grow
– AI-augmented platforms for faster testing, operations, and support
– Solutions engineered to handle complexity, pressure, and scale
If your team is facing a high-stakes challenge, we’d love to hear about it: get in touch with us.
Can Agents Replace Apps? We Put It to the Test
Can agents really replace traditional apps in enterprise environments?
We’ve been putting that question to the test in a multi-part LinkedIn series inspired by Satya Nadella’s “agent-first” vision—where business logic moves into intelligent agents that read, write, and update data across systems.
First, we unpack the shift and why it matters: thinner interfaces, direct interaction with data, and workflows that sound more like natural conversations than scripted clicks: → Read the vision →
The "app" becomes a conversation with an intelligent system, not a fixed interface.
Imagine:
“Show me my top five customers by revenue.
Email them all about our 2025 offer.”
Then, it just happens - no separate CRM, email client, or report builder.
Then we took our first step: API-first agents. We bound an LLM to OpenAPI-described endpoints, avoiding direct database access while keeping governance, observability, and policy enforcement intact. It worked well: ~89% success on standard operations. But it broke down whenever a required endpoint didn’t exist: See the results →
Next, we tested the other approach: SQL-first with Microsoft’s MCP Server. The agent had total access:every table, every operation was available immediately. But even simple tasks required excessive guidance and proved that raw access isn’t viable in production. SQL offered full access and flexibility, but required far more prompting and often lacked the business context to complete tasks correctly. The missing piece: structure and context. See what broke →
Now we’re moving into the next part: when neither approach is enough -LinkedIn post. What’s the right architecture when APIs are too rigid and SQL is too risky? What we learned: Neither approach works on its own. The real constraint isn’t the interface—it’s where business logic and semantics live.
The next post explores this Database Dilemma, along with real numbers and practical patterns for building agentic systems that are both powerful and safe.
Stay tuned.
Join Us at These Upcoming Events!
Kate Nyzhehorodtseva, Akvelon’s Director of New Business Development, will be attending:
- FinovateFall – The leading fintech event showcasing live demos, expert insights, and next-gen solutions for banking, payments, AI, and financial services.
September 8–10, 2025 · New York, NY, USA - AI & Big Data Expo Europe – A major European conference on AI and big data innovation, covering real-world applications, enterprise strategy, and the future of intelligent platforms.
September 24–25, 2025 · Amsterdam, Netherlands
To connect with Kate at these events or to set up a virtual meeting, reach out to her via direct messages on LinkedIn.
Insights from TechEx Santa Clara 2025
Earlier in the summer, our team joined over 7,000 tech leaders and innovators at TechEx Santa Clara. In our LinkedIn post, Ashley Pikle and Kate Nyzhehorodtseva share what stood out most as AI continues to reshape business.
Key takeaways from the event:
- AI agents are moving from theory to real workflows.
- Cybersecurity must change as attackers use AI.
- Infrastructure is becoming smarter and more autonomous.
- Leadership in the AI era means guiding people through change.
All of these trends point to one thing for businesses: AI must be built with both power and responsibility in mind.
Share your feedback and ideas about what you’d like to see on Akvelon’s LinkedIn by emailing info@akvelon.com!