AgileAlgo: Pushing the Limits of AI with AI

Tony Tay, Founder and CEO

Requirements analysis marks the first critical step in any software development endeavor. The requirements are collected and shared with development teams, often operating from different countries. But in this process, an offshore center functions like a mysterious ‘black box,’ where requirements go in, and code emerges.

“Could this black box process be digitalized?” probed Tony Tay, an IT visionary and former managing director for technology delivery public sector Southeast Asia, Accenture.

His extensive background, ranging from a CIO role in a regional brokerage to involvement in a Singaporean quasigovernmental initiative for global business expansion, sharpened his insights into software development.

Driven by this rich experience, Tony embarked on a journey to unravel how deep-tech and AI techniques could autonomously interpret requirements and execute coding tasks without depending on developers, enabling companies to do more with less resources. His goal was to empower companies to efficiently leverage the power of AI to address simpler use cases and leave data science teams to tackle more complex use cases.

He conceptualized this transformative approach as ‘natural language to code generation,’ which was developed into a full-fledged platform called AgileAlgo.

Tony combined agility and innovation with algorithms, placing AgileAlgo at the forefront of innovative algorithmic design, and giving birth to a concept that closely aligns with today’s generative AI. .

“We are using AI to create another AI today. Think of AgileAlgo as a smart workflow management software that seamlessly collects user requirements from every stakeholder in an agile manner and integrates that with a repository of codes. This fully integrated solution is the next big step to democratize access to AI technology,” says Tony, founder and CEO of AgileAlgo.

With AgileAlgo’s platform, business analysts or non-data science personnel can write requirements that can then be fed into an engine designed to interpret them and produce the desired results. This facilitates rapid prototyping, allowing organizations to quickly iterate and refine their solutions.

In the world of enterprise software, where even the smallest projects often involve seven-figure budgets, AgileAlgo sees a substantial opportunity to reduce costs and increase efficiency with an engine that can directly generate code from requirements.

How is AgileAlgo any different from no-code platforms? .

Most no-code and low-code platforms have a set interface where users begin by configuring various settings to initiate the development process. Although user-friendly, it does not prioritize the documentation of a user’s specific needs and business objectives before commencing software development.

In contrast, AgileAlgo emphasizes customization and prioritizes detailed requirements documentation as the starting point, making it a great digital alternative to a manual development process. It ensures a clear translation from business needs to technical execution, mitigating the risk of misinterpretation.

Impressively, it leverages graph neural networks to develop inference engines in less than five minutes and supports most AI algorithms, including big data analytics, NLP, computer vision and search.

The ability to convert natural language to code solves another pervasive issue in software engineering— requirements tracing.

Discrepancies that may arise in software development can be traced back to specific requirements to be adjusted or refined. This approach results in the generation of the right code, which can be changed again by simply updating the natural language inputs.

While some might draw parallels to ChatGPT or similar AI models, AgileAlgo stands out by employing a multiprompted approach.

ChatGPT may provide sample code that requires further refinement, but AgileAlgo aims to generate fully functional applications directly from natural language input without additional coding.

Enhancing Business Processes with Everyday AI

One big advantage AgileAlgo brings to the table is the implementation of smaller and more practical AI applications.

While larger AI-driven projects follow a standardized approach to development and execution, the challenge for most companies lies in developing and integrating everyday AI into their systems. This struggle is typically due to the complexity and resource-intensiveness of tailoring AI solutions to fit a business’s distinct, day-to-day operational needs.

AgileAlgo’s user-friendly platform easily addresses this gap.

“Similar to the evolution of user experience, once overlooked but now a critical component of digital project design, we envision a turning point where AI is embedded in every minute part of digital systems, enhancing their intelligence,” says Jonathan Ang, data scientist and co-founder of AgileAlgo. “This direction represents the future of technology.”

Integrating everyday AI is a major step forward, providing practical analysis tools to companies managing vast quantities of data. Traditional methods, which primarily handle numerical data, are no longer sufficient in a world where data comes in various formats.

AgileAlgo democratizes AI, making it feasible for businesses of all sizes to integrate intelligent automation into their daily operations without a large team of data scientists.

With the shortage of data scientists in the data management sector, organizations can allocate the complex AI use cases to their existing team and employ AgileAlgo’s platform to address the simpler ones, significantly expanding their capacity to implement AI without the need to increase specialized personnel. This is especially beneficial for SMEs who frequently face the challenge of losing their data scientists to larger tech companies offering better compensation.

With AgileAlgo, the efficiency of the existing team of data scientists can also be improved. As its strategy involves taking existing code and templating it within its platform repository, developers can use those templates later, instead of writing new code from scratch. This way, the overall software development timeline can be streamlined with each cycle.

Leading with Proven Success Stories

Numerous SMEs have used AgileAlgo’s platform to step up their digital systems with AI tools. For instance, an e-commerce business needed help with product discovery due to the complexity of accurately categorizing its wide range of products. The issue was in ensuring that products were organized in such a way that they could be easily found by customers searching the platform.

The client leveraged AgileAlgo to enhance their digital platform’s intelligence by implementing multiple recommendation engines. The engine was created in a single day. The AI-powered engines now recommend appropriate categories and subcategories for products based on their descriptions, titles and personal profile information. It streamlines the search process for the end customer, eliminating the need to manually determine the most fitting categories for products or navigate a vast list of options—a common issue in many of today’s e-commerce environments.

"workflow management software that seamlessly collects user requirements from every stakeholder in an agile manner and integrates that with a repository of codes. this fully integrated solution is the next big step to democratize access to ai technology"

Another success story features an HR firm specializing in peripheral HR processes. The firm uses a module called Quest for skill training. Setting up Quest involves specifying which skills the training is aimed at. A challenge arose from the extensive skill list and inconsistent terminologies used by different users. One user typed ‘automation’ while the other typed ‘RPA’ to indicate the same skill, causing discrepancies within the system.

To streamline this process, AgileAlgo developed a recommendation engine for the Quest within a couple of hours. This engine assisted users by suggesting relevant skills based on their input during setup, significantly upgrading the user experience.

Explaining the value brought in by AgileAlgo, Tony says, “Such a task would have been impractical with human effort alone, given the cost considerations. Often, companies allocate substantial funds to data scientists to tackle major projects that can take months to years. Our approach allows for tackling these everyday challenges without incurring high costs.”

Redefining AI with Talent, Technology and Global Aspirations

The company has come a long way since embarking on this journey three years ago, at a time when knowledge graph data science and graph neural networks were emerging fields.

Central to AgileAlgo’s success is Tony’s strategic decision to hire individuals with strong backgrounds in mathematics and statistics, which he deems the core of AI-related technologies. This approach bore fruit, notably with the recruitment of Ang, a mathematician whose expertise significantly enhanced AgileAlgo’s capabilities over the years.

Maintaining these foundational philosophies, Tony continues to assemble an expert team adept in front-end technologies such as the MEAN stack, comprising Mongo, Express.js, Angular, and Node.js. This blend of talent underscores AgileAlgo’s commitment to technological excellence.

Beyond the technical and data science team, AgileAlgo has also started recognizing the importance of its market presence and has intermittently augmented its sales and marketing divisions. By hiring seasoned marketing professionals, AgileAlgo aims to fortify its go-to-market strategy and broaden its operational horizons.

AgileAlgo has entered into a business combination agreement with a special purpose acquisition company (SPAC) listed on Nasdaq. Aimed at public listing, this move represents a strategic leap forward and, when the transaction is completed following receipt of applicable approvals, will propel its global aspirations.

With its proprietary technology, AgileAlgo seeks to assist organizations in navigating the complexities of data management and AI use cases while addressing the resource scarcity in the AI sector. Through these endeavors, AgileAlgo aspires to position Singapore prominently on the global stage as a net code exporter.