How we use AI at Pointr

by

Emin Sadiyev

18/06/2025

AI is evolving rapidly. Every week, we see new models with improved capabilities—better reasoning, smarter coding assistance, and more. Social media is filled with impressive AI demos, and companies are racing to integrate these tools into their workflows.

But with all this excitement, one big question remains: How can organizations actually use AI to boost productivity and efficiency? Many struggle to bridge the gap between AI’s potential and practical implementation.

At Pointr, we’ve gone beyond experimentation and developed real strategies to make AI work for us. Instead of chasing every new feature, we focus on workflows where AI delivers measurable benefits - just as we've done in our own AI-powered tool, MapScale®. In this post, we’ll share our approach to maximizing productivity with AI while keeping costs under control and maintaining quality standards.


Smarter AI Access with LibreChat

For teams to work efficiently with AI, they need access to the best models. But choosing the right way to provide that access isn’t always simple.

Many companies rely on subscription plans from AI providers like OpenAI, Anthropic, and Google. While this is convenient, it has drawbacks:

  • Lack of flexibility – Some models excel at specific tasks. Sticking to one provider means missing out on others’ strengths.

  • High costs – Subscriptions can be expensive, especially for large teams. A standard subscription model for a 100+ member engineering team can cost thousands of dollars per month, even if actual usage is low.

  • Limited adaptability – AI evolves fast, and locking into one provider makes it harder to switch to newer, better models.

To solve these issues, we use LibreChat, an open-source AI platform that lets us integrate multiple models in one place. Here’s how it helps:

  • Easy model switching – Our team can pick the best AI model for each task.

  • Significant cost savings – With API-based billing, we only pay for what we use. This has reduced costs dramatically—sometimes by an order of magnitude—while maintaining access to top-tier AI capabilities. 

  • Future-proofing – As new models emerge, we can integrate them quickly.

  • Smooth enterprise integration – LibreChat works with authentication tools like Active Directory, making it easy to use in corporate settings.

By using LibreChat, we’ve created a cost-effective, flexible, and future-ready AI solution for our teams.


Faster, Smarter Code Reviews with Qodo PR Agent

Code reviews are essential but time-consuming. To improve efficiency, we integrated Qodo PR Agent, an AI-powered assistant that helps us review code faster and more accurately.

With Qodo, we’ve seen:

  • Quicker issue detection – It catches bugs, race conditions, and other problems that might be missed.

  • Consistent coding standards – Ensures all engineers follow the same guidelines.

  • Less manual review time – Engineers spend more time on complex logic and less on routine checks.

One developer said, “Qodo caught a deadlock that we might have missed in manual review. It’s like having another senior engineer checking your work.” AI-driven code reviews have made our development process more efficient and reliable.


pexels-pixabay-373543AI-Powered Meeting Notes with Gemini

For a global team, keeping track of meetings is crucial. We use Gemini to automate meeting notes, which has transformed the way we document and share information.

Gemini provides:

  • Accurate, detailed notes – No more manual note-taking.

  • Automatic action item tracking – Follow-ups don’t get lost.

  • Searchable archives – Past meetings are easy to reference.

This is especially useful for remote teams, ensuring that everyone stays aligned—no matter where they are.


Boosting Developer Productivity with Cursor

After testing multiple AI coding assistants, we switched from GitHub Copilot to Cursor. This decision has significantly improved our developers’ workflows.

Benefits of Cursor include:

  • Smarter code suggestions – It adapts to our coding patterns.

  • Better understanding of our codebase – More intelligent auto-completions.

  • Improved refactoring assistance – Helps simplify complex code changes.

  • Faster development – Reduces time spent on routine coding tasks.

Many of our developers describe this as “vibe coding”—a state of deep focus where AI enhances speed without interrupting creative flow. Cursor has helped our team save 1–2 hours per day, freeing up time for bigger challenges.


AI-Enhanced Knowledge Sharing with Onyx

A major challenge with AI is that its knowledge is limited to what it was trained on. To fix this, we use Onyx, a tool that makes our company knowledge base searchable by AI.

With Onyx, our AI systems can pull real-time, relevant information from internal sources like documentation, Jira tickets, Slack messages, and Google Drive. This helps in:

  • Keeping AI responses accurate – AI always has the latest company knowledge.

  • Better customer support – AI can reference past cases to provide informed answers.

  • Faster decision-making – Employees can find past discussions and documents easily.

By integrating Onyx, we’ve turned AI into a reliable, up-to-date assistant for our teams.


AI-Driven Testing for Better Software Quality

We’re exploring AI-driven testing to improve software reliability. Early results show that AI can:

  • Generate comprehensive test cases – Including edge cases that humans might miss.

  • Adapt test scenarios dynamically – Reducing test maintenance effort.

  • Find vulnerabilities faster – AI-driven testing is helping us catch more issues early.

This initiative is still evolving, but it has the potential to revolutionize our quality assurance process.


pexels-googledeepmind-17483868-1Practical AI Adoption: Lessons from Our Journey

AI is evolving fast, but success isn’t about using every new tool—it’s about choosing the right ones for real problems. At Pointr, we’ve focused on practical AI integrations that:

  • Enhance productivity – From AI-powered coding to automated meeting notes.

  • Reduce costs – Flexible AI access prevents overspending.

  • Maintain high quality – AI-driven testing and reviews improve reliability.

By selecting AI tools that solve real challenges, we’ve seen tangible improvements across our teams. And as AI continues to advance, we’re committed to refining our approach to maximize its benefits.


The result? Our own AI mapping tool, MapScale®!

One of the major successes of the Pointr team over the last few years has been the launch of our AI mapping tool MapScale®. Born from a specific need from one client who came to us several years ago as they were manually creating maps for each of their stores in an image editor, MapScale® has since gone from strength to strength, transforming billions of square feet's worth of static floor plans into beautiful, interactive, 3D maps.

We've got a comprehensive breakdown of how MapScale® leverages AI to produce market-leading indoor maps here.

Intrigued about the latest with MapScale®? Our latest video showcasing v9.0 of Pointr MapsTM features a healthy dose of the latest MapScale® updates:

 

And of course, if you want to try MapScale® out for yourself, we have a free demo!

Launch MapScale Demo

 

by

Emin Sadiyev

Emin was formerly Pointr's VP of Engineering, and was integral in leading our global engineering teams towards an AI-first development mindset, meaning he's ideally placed to discuss the role that AI has played in the creation of Pointr's market-leading technology and solutions.

You may also like

Technology
How we use AI at Pointr
by

Emin Sadiyev

Read more
Workplace
LinkedIn's Path to $2.6M Savings and an Improved Employee Experience
by

Chris Schwartz

Read more
Technology
How Pointr Changed the Game with Indoor Blue Dot Technology
by

Daniel Murphey

Read more