Build & Train AI-Agents, Automate Transactional Work, Help Humans Be More Human Support Your KIM Strategy
Agentic AI In Social Housing
The potential for Social Housing in using LLMs alongside Machine Learning represents a whole new world of possibilities. The latest advanced reasoning LLM's are designed to interpret emotional cues from user interactions, including tone, language choice, and behavioural patterns, allowing them to generate empathetic responses tailored to the customer's emotional state; critically important in Social Housing where managing need & vulnerability is a huge aspect of what we do.
​
The ability to generate human like conversations through Advanced Natural Language Understanding, combined with the ability to reason, and demonstrate that reasoning through displaying the models "Chain Of Thought" open's up a huge range of possible applications within Social Housing. This advanced reasoning can be combined with the ability to create "chained" LLM Agents, each with their own role, instructions and separate knowledge bases, allows the creation of Agentic AI Teams (or "AI Crews"), enabling very technically accurate, human like responses, making assisted self service a reality.
​​
Building AI Agents
AI Agents can be internally facing & act as expert support to staff members on specific topical areas (such as Social Housing legislation) or can be externally facing and communicate with customers. They can also call functions & retrieve and update information from other systems, where they are permitted to do so.
​
Each Agent can be given it's own role, objectives, area of expertise & tone and behaviour. These can be simple or complex in nature, and the agent can be told to adjust it's responses based on user inputs & the emotional context of those inputs.
​
We can help you create your AI strategy to understand how Agentic AI can automate transactions & enable self service at scale, releasing staff to focus on more human work.
Building Knowledge Bases
Knowledge bases can be created based on policies & procedures, or actual examples of the outputs that you want the Agent to create, these knowledge bases train the model through RAG (Retrieval Augmented Generation- the process of optimising the output of a large language model, so it references specific information, outside of its original training data sources before generating a response). In effect, we are giving the model organisational & sector specific information to customise its responses, based on it's intended role. ​
​
The creation of agents with knowledge bases can support your internal Knowledge & Information Management (KIM) strategy, helping you to collect, store, share, and use data and knowledge. The housing ombudsman described KIM is "the closest thing the sector could get to a silver bullet" for addressing the issues in social housing.
Assigning Guard Rails
Guardrails are security components that you can attach to agents. A guardrail overrides an agent’s output when it detects sensitive or inappropriate content in the agent’s input or output & provides a layer of defence from bad actors attempting to "jailbreak" Agents. They can also be used to protect sensitive information. ​
Defining Agent Routings
Agent routing is the ability to reference other agents from the instruction of an agent; there can be a general customer service agent which acts as the "parent" agent but has instructions to hand off the request to a "child" agent depending on the subject matter- for example an agent specifically trained in tenancy sustainment & arrears. By creating a team of agents, giving specific roles & instructions and attaching very specific knowledge bases, the quality & technical accuracy of the responses can be greatly improved. ​​
Building & Deploying Your Agent Teams
Your AI Strategy should define the goals you want to achieve, both in terms of AI enabled insights, through Machine Learning & Deep Learning and in terms of transactional & knowledge based LLMs.
​
We can help you define your AI Strategy, Roadmap & Implementation plan, and create the data layer, knowledge bases, and AI agents & teams to deliver this strategy, and build and deploy your AI Agents to harness the benefits of AI & drive better customer and operational outcomes, whilst reducing your overall operating costs.
​
Benefits
Improve The Customer Experience
Long call wait times, siloed service delivery requiring call backs and frequent call transfers to other departments frustrate customers & build in failure demand; if it's just too much hassle to re-arrange that Gas Servicing appointment, customers won't bother. AI agents, blended with human agents, can deliver a much better, no silo customer experience, 24x7.
Multi Sentiment Analysis
The new generation of reasoning models, enable you to do multi-sentiment analysis, understanding basic sentiment (satisfaction), but other predictions such as the likelihood of a complaint, or vulnerability & safeguarding sentiment markers & can adjust the conversation accordingly, offering help and support.
Generate Significant Cost Savings
Deploying AI Agents, combined with opening up new social channels of engagement with the Engage-AI solution can slash the operating cost of your contact centre. By opening up digital channels & moving to an AI enabled self service model, staff can be released to focus on human work and additional financial headroom generated.
Increase Productivity -Knowledge Based Co-Workers
Create Agents & knowledge bases to support specific roles within your organisation & make staff more productive. Input policies, specific examples and train the agent in the expected outputs and increase efficiency across the workforce. Reduce admin time & maximise time spent with customers.