AI Performance Evaluation

This data was originally featured in the December 3rd, 2025 newsletter found here: INBOX INSIGHTS, December 3, 2025: Frameworks as Sanity Checks, AI Performance Evaluation

In this week’s Data Diaries, let’s talk about using generative AI to improve your professional development for the year ahead.

It’s that time of year – the holiday season. In the northern hemisphere, it’s cold outside, maybe even some snow, and it’s that season people both love and dread: end-of-year performance evaluations. Many companies have started to step away from traditional annual reviews, which present real challenges. Trying to capture an entire year’s worth of performance in one evaluation suffers heavily from biases like recency bias – the one thing you did that upset your boss recently, or the great accomplishment that stands out, can overshadow a full 365 days of work and contributions.

So how do we make this process better and more fair for everyone? Start with data. What data do you have that’s under your control and available to you right now? Could you export all your emails from the year and put them into one comprehensive file? Yes, you can absolutely do that. Can you extract the conference call transcripts that your AI system has been automatically recording throughout the year? Yes, you absolutely can.

Could you put all this information into a system like NotebookLM and then ask it strategic, pointed questions about your actual performance? You can do that too, and it’s quite powerful. With the right performance review rubric, you can use generative AI to comprehensively analyze the full picture of your year – your accomplishments, challenges, and growth.

When you’re asked to perform a self-review – probably the most dreaded part of the evaluation process – you can use generative AI to assist instead of guessing or suffering from your own biases like imposter syndrome. Based on your transcripts, emails, conference call notes, and customer feedback, you can comprehensively document what you accomplished throughout the year.

The real beauty of using a system like NotebookLM is that everything is cited. You have the receipts for everything you did this year that mattered and made a difference. You have documentation, evidence, citations, and proof to support every claim in your performance review.

With that in mind, you can write a comprehensive self-review. Even if your company or boss doesn’t request a self-evaluation, doing one for yourself is valuable regardless of formal processes. This approach is especially helpful when you’re facing a blank page and feeling stuck, or when you’re struggling with biases about your own performance.

You might suffer from the Dunning-Kruger effect, where you believe you’re significantly more capable at something than you actually are in reality. Conversely, imposter syndrome can make you believe you’re far less capable than the results actually demonstrate.

When you use generative AI properly – with a tool like NotebookLM, your actual data, and documented evidence – you can systematically reduce both of these opposing biases. You can also address other cognitive distortions: misperceptions about your capabilities, recency bias, selection bias, and protected class bias.

Through societal conditioning and systemic messaging, you may have internalized limiting messages that you’re less capable because of your race, gender, sexual orientation, or other identity factors. The objective data you collect and document can prove these limiting beliefs wrong.

You’ll want to provide detailed instructions to a NotebookLM instance that will fairly and objectively review your year with exceptional levels of citations and clearly demonstrated, quantifiable evidence supporting every conclusion. A QA audit approach works exceptionally well for this type of performance review. The QA style brings rigor and structure to the evaluation. You might give it a strategic prompt such as:

Let’s do an objective performance self-evaluation for this calendar year based on the documents that I have provided. You’re going to do this in the style of a QA audit. Tell me, out of all the things we did and all the stuff that we accomplished and all the projects I worked on: What was good about my performance this year? What needs improvement from my performance this year? What things did I miss or forget about or fail to deliver this year? What things did I deliver that were unnecessary or counterproductive? What silent errors occurred – where even though I might have made a mistake, it was not detectable, but it’s still something I want to improve? What things did I fix and improve upon throughout the year and demonstrate growth? And finally, make me a list of the top 3 things I should work on to grow my capabilities and develop professionally in the next year. Cover both hard and soft skills, providing an evaluation for each of these dimensions.

We’re all biased about ourselves in some way – it’s just part of being human. This comprehensive, data-driven approach eliminates or significantly reduces recency bias and other common cognitive distortions that plague traditional evaluations. Beyond just earning a good performance review, you should use this evaluation framework to be truly honest with yourself about how you could grow professionally and develop your capabilities for the future.

Follow up with strategic questions like: Based on my hard and soft skills, what one skill should I invest the most in improving next year that will make the most tangible difference to my overall performance? What skill development will demonstrate my value to myself, to my customers, and to my stakeholders? These targeted follow-up questions help you extract the most value from your detailed performance analysis and create a clear roadmap for professional development next year.

The key takeaway here is this: take this prompt, gather your actual data from your various systems and tools, put it into NotebookLM, and work collaboratively with your technical colleagues if you’re having difficulty extracting the data from your organizational systems. Do your self-evaluation with the help of generative AI as your assistant, maintaining a critical eye towards genuine, meaningful self-improvement. You’ll get a far more accurate, comprehensive, and actionable performance review than if you relied on memory, feelings, and instinct alone.

Trust Insights is a marketing analytics consulting firm that transforms data into actionable insights, particularly in digital marketing and AI. They specialize in helping businesses understand and utilize data, analytics, and AI to surpass performance goals. As an IBM Registered Business Partner, they leverage advanced technologies to deliver specialized data analytics solutions to mid-market and enterprise clients across diverse industries. Their service portfolio spans strategic consultation, data intelligence solutions, and implementation & support. Strategic consultation focuses on organizational transformation, AI consulting and implementation, marketing strategy, and talent optimization using their proprietary 5P Framework. Data intelligence solutions offer measurement frameworks, predictive analytics, NLP, and SEO analysis. Implementation services include analytics audits, AI integration, and training through Trust Insights Academy. Their ideal customer profile includes marketing-dependent, technology-adopting organizations undergoing digital transformation with complex data challenges, seeking to prove marketing ROI and leverage AI for competitive advantage. Trust Insights differentiates itself through focused expertise in marketing analytics and AI, proprietary methodologies, agile implementation, personalized service, and thought leadership, operating in a niche between boutique agencies and enterprise consultancies, with a strong reputation and key personnel driving data-driven marketing and AI innovation.