Topher Williams

This was a complicated project due to the newness of government requirements and regulations in the cannabis space, and the general lack of technological sophistication of most commercial growers at the time.

Working with the company's CEO, I helped his team build and manage a development backlog, refine feature prioritization, interview clients, and oversaw the development project for more than a year.

The project was technically tricky in that it covered a vast array of technologies: cloud backend and web application, a native mobile app with RFID reader integration, fixed infrastructure RFID readers, and on-site middleware. In addition to the wide variety of technologies in play, the system needed to meet various government (and sovereign nation) reporting and auditing requirements, most of which were not well defined as we were developing the system.

The MomCo app was a mobile app (native iOS and Android apps) that used geolocation services to help moms find other moms with similarly aged children near. The app also promoted company hosted playdates and other events around the US, private mom groups, and a host of other features.

The app was featured on the Today show and attracted more than 50,000 moms to the platform. The engagement and adoption that the app was able to generate attracted $300,000 in seed funding during my tenure with the company.

With the ability to get moms (a highly valuable demographic) to attend events and provide feedback on products and services that were available to try at these events, we started work on developing features to allow the app to be used as a virtual consumer panel. We did this so we could sell access to selected collections of moms to brands that wanted to sell to that demographic. This was an exciting expansion opportunity for the platform.

The client is an online retailer looking to optimize the allocation of their capital across hundreds of thousands of products that they could potentially sell. Adding another level of complexity, the client was also selling established products on Amazon, which meant there was dynamic competition, pricing pressure, and other factors present that made demand and price forecasting particularly tricky.

Working with the client, we devised an ideal workflow that would enable maximum efficiency and automation. The workflow would use algorithms to identify potential products from Amazon's 450 million item product catalog that might be good performers for the client to sell.

The algorithms would add brands to a sales pipeline tool where workers would process the leads and attempt to establish relationships with the manufacturers and obtain pricing information (and other logistic information). The workers would then submit that data to another software solution that would perform an in-depth financial analysis of each product, including leveraging several machine learning algorithms, to identify the optimal inventory investment mix that the business could act upon. This efficiency allowed the company to scan hundreds of thousands of products.

The client was interested in building a SaaS product to replace in-house financial analysis and reporting spreadsheets and other AIS tools they used with a more automated and integrated web solution that could also be sold to other firms.

The client being a full-service accounting firm had no in-house technical resources or any experience bringing technical products to market. For this project, I worked with the client as if I were an in-house CTO.

In my role as fractional CTO, I worked with the client to capture and refine their product vision and requirements, decomposed those product requirements in to user stories, and defined the technology stack and system architecture for the project. Once this up-front work was completed, I managed a vendor selection process to qualify on- and off-shore development firms and guided the client through the partner decision making process. Once the development firm was selected, I worked with them constantly throughout the project to both manage the agile development process with them, serving as the product owner, but also in managing user acceptance testing and other responsibilities of the client, while transparently communicating project status, issues, and so forth to the client.

For this project, I worked on defining what the MVP of a voice analytics product should look like.

I presented an in-depth analysis of the marketplace, from the perspective of who the players are, what business models are in play, what technology relationships exist, and what each competitor's offering looks like.

I worked with several existing and potential customers of the clients to identify their pain points, concept test potential product features and business models, and really try to understand how these technologies could create value in their organizations.

With this information, I developed an MVP concept, including which features should (or should not) be a part of the MVP (and what a feature roadmap might look like—similar to a PRD), as well as a few business model options and how each of the models might perform (considering various development cost scenarios combined with different operational cost and pricing models).

After receiving buy-in from the client's stakeholders, I worked through the MVP definition and created user stories to break down and describe each of the MVP's features, and technical evaluation and selection criteria for the technology partners that I had previously identified.

As Q.ai's head of product, I led a team of designers, researchers, product managers, and customer support representatives to apply data-driven experimentation and decision-making, resulting in more than a 10% monthly growth rate in funded accounts and churn reduction to less than 3% annually. Additionally, I implemented an OKR process and developed processes to systematize innovation across the organization. With ownership of the product backlog, I drove the success of Q.ai and helped it grow under Forbes' umbrella.

The client is a sales and marketing agency that uses several tools and spreadsheets to engage in outbound email marketing and appointment-setting activities. I worked with the client and development teams to build an easy-to-use web application that automated several labor-intensive tasks like sourcing, importing, validating, enriching contact information, writing email copies, organizing contacts based on campaign performance, and reporting). Cost reductions were achieved by automating as many processes as possible by integrating with various data brokers in a cost-effective manner, leveraging AI for copywriting, and building a reporting platform to serve the business and its clients.

Led a team of engineers to design an RFID-based system providing cryptographically secure, supply chain asset authenticity validation. The system is comprised of a web application, server applications, and embedded hardware applications. I worked with the US Air Force, Lockheed Martin, and other stakeholders to define system requirements that meet cyber security and other DoD standards. We delivered a system that allowed the USAF and their suppliers to verify the authenticity of materials and components and to trace each item's lifecycle through manufacturing and repair processes.

Spearheaded the development and implementation of an innovative, AI-powered mobile fitness game, leveraging OpenAI's ChatGPT to personify various in-game characters, enhancing interactivity and user engagement.

Utilized prompt engineering to effectively train ChatGPT, creating diverse AI personalities, each with specialized knowledge, providing a unique and engaging gaming experience.

Actively collaborated with a cross-functional team to ensure the AI integration complemented the game design while optimizing the overall user experience.

Led rigorous testing and debugging procedures, ensuring optimal AI performance and contributing to the delivery of a high-quality, robust gaming product.