Greater Pittsfield Area
A lot of organizations spend a ton of money on new technology and then watch it just... sit there. The team doesn't use it the way it was meant to be used, the investment starts to feel like a waste, and nobody really knows why it isn't working. I've seen this problem up close for over two decades. At Apple, I spent more than 12 years helping people actually connect with technology in a way that made sense to them. I wasn't just facilitating software. I was figuring out why smart people were stuck, finding the knowledge gap between the tool and the person using it, and closing it. I did that in small groups, and at scale of 200 people. That experience taught me something most consultants miss: the technology is rarely the problem. The system around it is. Today I use Lean methodology to help organizations clean up that system. I find the waste, fix the flow, and make sure the technology your team is using actually reflects what your business stands for. I also bring an ethics-certified lens to the work, because in today's world, how you build your systems matters just as much as what they do. I use Business Process improvements to get granular and find the gain points and pain points. If your technology feels like it should be doing more for your business than it is, that's usually a sign something in the process needs to change. That's exactly what I do. Let's connect.
I am currently working in a contracted role focused on multimedia AI generation and testing, supporting structured AI production and evaluation workflows. Client Unknown. The work involves completing and reviewing AI generated multimedia tasks through a dedicated platform while following clear instructions and scoring guidelines for each assignment. Each task is structured and typically takes around 3 to 4 hours to complete Work is evaluated based on accuracy, consistency, and adherence to detailed rubrics and measurements. Tasks are managed through a controlled system with defined rules and workflows Time tracking and communication tools are used to ensure accountability and clarity Guidelines and scoring systems are updated regularly as project needs evolve. The environment is fast paced and requires attention to detail, discipline in following process, and consistency in output.
In this role as a contract AI Data Architect, I create structured data solutions to support a global initiative. I apply domain expertise to build realistic, high-quality data inputs that help drive decision-making and ensure diverse, enterprise-ready outcomes.
Quality Assurance Auditor [Contract] Dates: November 2025 – January 2026 Promoted to QA Auditor based on high-quality performance and consistency in multimodal data production. Evaluate AI training tasks submitted by experts using a four-point quality scale to ensure strict compliance with client rubrics. Support the team by leading with quality and by example. Identify data quality gaps and guideline violations across complex multimodal tasks. Provide constructive written feedback to contributors, detailing technical inaccuracies and necessary corrections to improve model reliability. Collaborate with leads to resolve edge cases and maintain alignment as project instructions evolve. Maintain strict confidentiality of audit materials and adhere to secure communication protocols. Role 2: Multimedia Data Contributor [Contract] Dates: August 2025 – October 2025 Produced high-quality multimodal training data to support advanced AI research initiatives. Delivered structured spoken descriptions of complex image sets for large-scale AI datasets. Followed precise linguistic and technical guidelines to ensure 100% annotation accuracy. Maintained high throughput and met aggressive turnaround targets in a remote production environment.
I tested and was trained in the AI system to see how safe and reliable it was. My job was to act like a real user and try to push the system into making mistakes, especially in serious situations involving mental health. I looked for places where the AI might respond the wrong way or miss warning signs. This helped show where the system could be improved to better protect users.