What we do

Client wants to utilize GenAI in transformation or customer-facing solutions.

Frontier are pioneers in using GenAI since 2014 when we developed a décor language model to generate art descriptions using computer vision. We offer the full range of GenAI services, from model training to fine-tuning to prompt engineering and deployment.

All of our recent projects involve GenAI or LLMs e.g. a model to generate business summaries for a business directory service.

Client wants a proof-of-concept or an entire scalable solution.

We offer a range of software services using modern stacks and Agile methods. To accelerate development, we built our own microservices framework on AWS, which we call the Frontier AI Accelerator Kit. We can deploy a skeletal app at the push of a button and quickly start trying out ideas or even launching an MVP.

For an online travel agent (OTA), we built a demo product to explore a novel OTA concept.

Client wants a rapid validation of an AI-related idea: can this be done? What are the limitations, issues, caveats, challenges etc?

We offer a rapid validation service that can involve digging into core assumptions and quickly identifying the critical components and their validity. This can involve both paper- and code-based designs and tests.

For a wellness company we indexed their existing knowledge-based (KB) for use with Large Language Models (LLMs) to identify potential weaknesses in the KB for 'virtual practitioner' use cases.

Client wants to understand how and where to apply AI to which part of their business stack, anywhere from immediate to a 3-5 year 'Horizon 3' timeframe.

Our forte, using intense research via our holistic AI method is the discovery of priority use cases to drive real value. We can explore any aspect of organizational innovation, from customer-facing products to value-engineering cost optimizations.

For an online hospitality application, we identified a number of use cases for the use of Generative AI (e.g. in assisting with venue summaries or user recommendations).

Client has a hard or specific technical problem, such as how to improve driver safety for fleet management, minimize carbon footprint across a whole sector, etc.

We can identify and, if necessary, invent core IP for strategic programs right down to the algorithmic level. The solution can involve identifying patent strategies.

For a major fintech, we built a novel computer-vision pipeline to help with OCR document onboarding without the helpful constraints of a mobile app.

Client wants to assess general AI readiness for the org or for a specific range of problems, such as how to transform sales to an AI-first sales organization.

We can assess operations and data stacks in relation to AI transformation. This can include maturity ranking for various capabilities, such as data governance and quality. It can include design thinking to identify core end-user problems etc.

For a large global e-commerce platform provider we identified the readiness of their solution architecture in light of proposed AI trends that would impact requirements from their major retail clients.

Client wants to understand what AI-first transformation means for their business.

Our innovation strategy team assesses the core nature of the business and its sector in relation to AI-first opportunities. This includes broad strategic innovation.

For a wellness company, we proposed how corporate wellness might be re-imagined in light of AI-first possibilities, especially Generative AI.

Client wants to understand what AI alignment issues they should be aware of for a given set of AI proposed solutions or, more generally, for their expected use cases.

We can assess AI alignment and identify risks along with solutions aligned with appropriate ethical and legal frameworks.

For a major fintech, we identified how bias in the use of opaque AI models might be overcome using a range of mitigation and alignment tools, some of which we invented.