At Being Guided we want to help organisations of all shapes and sizes to achieve more value from less IT spend on Salesforce solutions. Our venture with Supercog helps our Salesforce designers and developers work faster, with less errors. This means automating and guiding tasks using the power of AI Agents - Large Language Model (LLM) programs connected to live systems and data.
Supercog is on a journey to work with enterprises and system integrators who want to achieve new and improve existing integrations with Salesforce. As illustrated above, this means applying the Supercog AI Agent Platform to create custom and productised Salesforce Integrations, built on industry standard Apex code and Lightning Web Components (LWC).
With Supercog we are delivering a new AI concept. We recognise the reality of our prospective customers who have never heard of Supercog (or Being Guided) and may have a wary view of the real business value, risks and hype surrounding AI.
So, we want to emphasise a key question here that we will answer:
What is the cost of NOT implementing Supercog for Salesforce Integration?
Three Design Principles, Applied
We show how organisations can engage early in something that starts out as more of a concept, and less of a product, yet delivers value early and often. This means following three Design Principles:
01. Meaningful Journey. Purposeful stages with a clear end game.
As illustrated above, we are moving from 'quick wins' with Salesforce Integrations in the UK NHS, to a more sophisticated outcome for Application Modernization. This applies across many industries and organisations. It starts with early wins and gets deeper through Design Thinking and Mutual Value Discovery workshops.
02. Fierce Reduction. Eliminate everything you can. Less is more.
At each stage of the Meaningful Journey we are focused on creating value for our customer, by simplifying processes and tasks. This is challenging business-as-usual through Design Thinking and enabling Empathy Mapping with all stakeholders.
03. Progressive Disclosure. Prove value created, early and often.
What's crucial with Progressive Disclosure is a need to both visualise and measure the value created at each stage of the Meaningful Journey. This is validing the power of Fierce Reduction to realise a 'less is more' outcome in the digital innovation.
A Compelling ROI Model
As with any new technology, for our early adopters of Supercog we must answer a simple question:
What is the cost of NOT implementing Supercog for Salesforce Integration?
Firstly, let's look at the Return On Investment (ROI) Model - ageneral formula:
ROI = (Cost of Investment / Net Profit)×100%
To adapt this formula for an As-Is vs. To-Be comparison, consider:
Net Profit: This will be the difference in profits between the Future State (To-Be) and the Current State (As-Is).
Cost of Investment: This is the cost incurred to move from the Current State (As-Is) to the Future State (To-Be).
Given the above considerations, the formula becomes:
ROI = (ProfitTo−Be − ProfitAs−Is / Cost of Transition) × 100%
Where:
Profit To-Be = Profit or (benefit) in Future State
Profit As-Is = Profit (or benefit) in Current State
Cost of Transition = Cost to move from As-Is to To-Be
Note: If you're measuring benefits other than strict monetary profits, such as time saved or other intangible benefits, ensure you can convert these benefits into a monetary value for this to be valid.
To calculate the Return On Investment (ROI) from Digital Innovation with the specified inputs, we can formulate several equations. Let's define the variables first:
BVAs-Is = Current State (As-Is) Business Value generated per annum without Digital Innovation.
BVTo-Be = Future State (To-Be) Business Value generated per annum after investing in Digital Innovation.
CDI = Cost of Digital Innovation as a Recurring Annual Subscription.
ROI = Return on Investment as a ratio relative to the Cost of Digital Innovation.
CoD = Cost of Delay per day when not investing in Digital Innovation.
CoDN = Cost of Doing Nothing per day when not investing in Digital Innovation.
CoDday = Cost of Delay per day when not investing in Digital Innovation.
CoDNday = Cost of Doing Nothing per day when not investing in Digital Innovation.
Calculating ROI from Digital Innovation:
Net_Gain - BVTo-Be - BVAs-Is
Calculating ROI:
ROI - Net_Gain - CDI / CDI
The ROI is expressed as a ratio. Multiply by 100 to get a percentage.
Cost of Delay (CoD):
This represents the loss per day by delaying digital innovation. Assume the delay starts from the beginning of the year and goes on for d days:
CoD = BVTo-Be - BV As-Is (d x CoDday) - CDI / CDI
Cost of Doing Nothing (CoDN):
This is the loss per day for not implementing the innovation. Similarly, for d days:
CoDN = BVAs-Is - (d x CoDNday) - CDI / CDI
Supercog Technology Overview
As the developer's digital assistant (Copilot), Supercog can help with a large variety of online tasks. Unlike widely-available AI chat programs such as Chat GPT-4.o, Supercog safely connects to your real systems and real data, and helps developers get real work done, faster.
Supercog Agents are semi-autonomous programs that follow simple English instructions. Each Agent is powered by a Large Language Model (LLM) which gives it the ability to understand language, make plans and execute tasks, faster.
A Supercog Agent is composed of three parts:
Model. The LLM Model powers the Agent. You can choose amongst many different LLMs, including GPT3/4/4.o/4.o-mini, Anthropic Claude 3, and Meta LLama3.x.
Instructions. The Instructions describe the work that the developer would like the Agent to perform. If you just want to experiment it is not necessary to write any Instructions. But when you want to devote an Agent to a specific task, then the developer will want to explain that task in the Instructions.
Tools. You enable agents to perform work by giving them tools. Without any tools an Agent is limited to the capabilities of the underlying LLM.
Supercog Demos:
https://www.youtube.com/@supercog-ai
Supercog Works with Systems
Supercog is great for creating new Generative AI Apps and delivering legacy IT integrations. Building applications with code takes a lot of work, while prototyping with ChatGPT is limited to writing prompts and document analysis.
With Supercog developers can create a whole application as an Agent, including multi-step flows, testing different LLMs, and mixing lots of different actions including:
Text generation and code generation.
Document and image analysis.
Pulling information from the Web.
Text-to-SQL.
File and data processing.
Quickly accessing APIs.
Leverage Supercog's natural language skills and ever-growing training data to help developers explore systems at work:
Ask questions in natural language about any data or schemas from your database or data warehouse.
Ask for explanations of code. Supercog can understand (and write) even esoteric languages like Salesforce Apex.
Ask Supercog to introspect schemas in your CRM system, like "please explain the custom fields on our Salesforce Opportunity Object".
Supercog can also easily synthesise system-specific queries like Salesforce SOQL and JIRA JQL. Quickly explore systems with queries like "show me how many JIRA issues were closed in the APP project this week".
Supercog has a smart data processing layer that allows the LLM to orchestrate functions that can query, transform, and insert thousands of records efficiently. For example:
Please download all the Leads from Salesforce and save them to a Snowflake table.
Data transformation and synchronisation are similarly easy:
Read the excel file new_leads.xlsx, then map the columns of the file to the columns from the Salesforce Lead object. Include the "LeadSource" field and set it to "Supercog".
Then upsert those records into Salesforce Lead objects matching on the "Email" field.
Beyond just interactive tasks, Agents can be configured to run autonomously in reaction to different events. Here's a real world example:
Whenever a new user record appears in our app database, create an associated Contact record in Salesforce. Include Web research results based on the domain of the email address. Every morning, summarise the updates to any JIRA Tickets in a project and post the summary to Slack. Download new Load records from Salesforce, save them into a Google Sheet, and email the sheet to my SDR team.