MAGIC’s custom AI agents automate business processes that are:
– Intellectual but non-creative
– High value and repetitive
The AI agents learn from your real-life examples, act autonomously with humans in the loop, and follow your rules and policies.
| MAGIC Chief of Staff | User | |
| breaks down processes into tasks, assigns each to the right specialist AI agent and synthesizes responses | researches, analyzes, reports >>>>>>>>>>>>>>>>>>>>>>>>>
teaches, supervises, reviews <<<<<<<<<<<<<<<<<<<<<<<<< |
collaborates with the MAGIC Chief of Staff AI agent via a novel UI |
Each MAGIC AI agent is made up of:
LLM a Large Language Model chosen to best suit the agent’s domain of expertise
Tools a variety of private and public tools, including APIs, to enhance capabilities and integrate with private systems
Data public and private data in secure, isolated environments, to provide insights whilst maintaining confidentiality
MAGIC is built on concepts of multi-agent collaboration and human-AI fusion intelligence.
AI agents, directed by an AI Chief of Staff, work together on complex tasks and learn from human input, making solutions scalable.
AI agents are most effective when applied to tasks that are:
High-value
Select processes that are critical to your business outcomes and offer a strong return on automation in terms of revenue and costs.
Intellectual
Choose tasks that are currently completed by trained human experts. AI agents can distill expert knowledge to automate and scale processes.
Non-creative
AI agents thrive on clear objectives. Tasks should have easily articulated outcomes, rather than open-ended goals.
Repetitive
AI agents learn continuously. The more frequently a task occurs, the more capable the agent becomes and the greater the value to the business.
We assess your use cases to determine feasibility and impact, and provide a tailored proposal with timeline and pricing.
We develop your custom AI agents, working with you to integrate with internal databases and systems, as well as to test and validate results.
Post delivery, we provide support to ensure long-term success.
Asset management: synthesize research papers, map out ‘street view’ on investment topics, performance reporting
Commercial loan underwriting: loan application review and feedback, credit paper drafting, due diligence review
Private equity: sector mapping and deal reviews
Recruitment: CV screening, bespoke interview questions, employee onboarding
Marketing and sales: competitor analysis and lead generation
Operations: project management, customer onboarding, metrics reporting,
standard operating procedure documentation