a21.ai heralds the emergence of right-sized GenAI solutions for SMEs

Every enterprise leader strives to enhance productivity and cut costs by leveraging GenAI. However, small and mid-sized enterprises (SMEs) rarely have enough budget to experiment their way to success. They need GenAI solutions that cater to specific business needs, are compatible with their existing tech stack, and deliver quick returns on investment (ROI).

a21.ai is trying to address those needs using custom GenAI solutions embedded with relevant tech and functional know-how plus responsible and secure AI. It is working toward carving a niche by targeting SMEs in the crowded AI landscape.

Building customizable accelerators for rapid development and deployment using SLMs rather than LLMs

a21.ai is currently focused on developing and deploying three accelerators: intelligent data processing (IDP), synthetic data generator, and enterprise-grade GenAI chat. These accelerators use a combination of traditional AI, machine learning, and small language models (SLM) to implement GenAI tech.

SLMs are cheaper than large language models (LLM) due to their smaller model size and fewer parameters, resulting in decreased token and energy usage. While their capabilities are limited, SLMs can help SMEs deliver faster ROI with minimal investment by automating a set number of tasks. Moreover, a21.ai builds custom apps on top of these accelerators to meet their clients’ needs. These apps are designed to be scalable, ensuring data confidentiality and compliance with regulations, and are regularly updated to align with evolving business requirements and technological advancements. This is essential for SMEs that have no access to large legal and compliance departments available to larger enterprises.

Two POCs using SLMs—one for personalization of online clothes and the other for policy generation for an Indian health insurer

From the handful of proofs of concept (POC), a21.ai demonstrated two that they built for mid-sized clients:

• The first POC was about a synthetic data generator accelerator—a21.SYNTH. This GenAI solution offers a virtual try-on feature for an ecommerce company, focusing on delivering accurate garment fitting and a uniform user experience. It uses sophisticated algorithms for body size and type classification, offering virtual garment-fitting mirrors to ensure the precision and personalization that online shoppers demand. The solution helped train AI models, eliminating the need to buy databases or manually collect data, thus lowering the cost of developing a model for a personalized customer experience.
• The second POC was about an IDP accelerator—a21.iDOC—utilized as a contract analysis and underwriting workbench for a large Indian health insurer. a21.ai initially used traditional methods to extract data, such as text extraction and image analysis tools, and then implemented SLM-based guidelines to generate code for policy generation. This IDP tool helped automate policy generation, freeing up valuable time and increasing productivity for employers.

A single system for information extraction and inference tasks targeted at SMEs reduces the need for multiple models

a21.ai aims to be a one-stop solution for SMEs, kick-starting their GenAI journey by supporting them with architecture, structure, methodology, and best practices. This involves preparing the data (including data orchestration, cleansing, and extraction), building full-stack GenAI apps, and keeping them compliant and up-to-date.

a21.ai’s integrated application engine promises to streamline processes by combining information extraction and inference tasks into a single system. This approach eliminates redundancies and enhances efficiency by avoiding the need for multiple systems and models. This integrated system can potentially refine operational workflows for SMEs seeking to minimize their investments in multiple models.

Pricing strategy to minimize initial cost and demonstrate rapid value

a21.ai requires a minimum upfront investment from clients to develop POCs for them. Once the clients are satisfied, they can transition to larger enterprise-level engagements. The reduced upfront investment is a smart pricing strategy to attract clients, as it gives SMEs (with restricted R&D resources) an opportunity to experiment with and understand GenAI solutions as well as their potential benefits and limitations.

a21.ai’s overall approach, entailing customized accelerators, integrated systems to streamline SME operations, and reduced upfront investment that breaks down entry barriers, could be a game changer for SMEs.

The Bottom Line: SME leaders must harness GenAI too. Firms such as A21.ai offer a low-cost, low-risk start.

SMEs must harness the benefits of GenAI to compete with larger rivals, but they have to keep costs and risks low. Exploring solutions from new companies such as a21.ai may be an important first step to starting their GenAI journeys.

Furthermore, SME leaders should welcome the opportunity but must demand domain and industry-specific solutions to take full advantage.