The AI Ecosystem
Generative AI (GenAI) applications are transforming the future of work with newer capabilities in reasoning, creation, and creativity. These advancements are driving enterprises to reconstruct their tech stacks to strategically generate significant value for their organizations. The evolution of the Enterprise-Grade Gen AI stack with more and more builders converging around infrastructure, tooling, and approach is giving enterprise buyers more confidence and clarity in what to buy.
Among the emerging building blocks of the Gen AI Stack, Lyzr sits in the Tooling layer which interacts with LLMs and helps build Generative AI applications.
Lyzr’s Low-Code Agent Framework
Early AI platforms like Langchain, LlamaIndex, and Crew AI focused on helping technical AI/ML developers build highly customized applications. Lyzr, however, offers a more accessible solution—a low-code platform that empowers both developers and non-technical professionals to design and deploy GenAI agents effortlessly.
Lyzr is the WordPress of agent frameworks, providing a low-code platform for anyone to build and deploy agents at scale. With Lyzr you can build chat agents, knowledge search, data analysis, RAG-powered apps, and multi-agent workflow automation with minimal effort.
The best part is that you can get locally deployable SDKs and private APIs to run the agents on your AWS cloud, thus eliminating all concerns related to data privacy, compliance, and latency.
Deploying Lyzr Agents on AWS
Enterprises can access Lyzr’s multi-agent framework in a private environment on AWS through two services: Amazon Bedrock and Amazon SageMaker. To access you will need to be a registered user.
About Amazon Bedrock
Amazon Bedrock is a fully managed service that provides a single API to access and utilize various high-performing foundation models (FMs) from leading AI companies. It offers a broad set of capabilities to build generative AI applications with security, privacy, and responsible AI practices.
Using Amazon Bedrock, you can build custom agents with Lyzr using techniques such as fine-tuning and retrieval-augmented generation (RAG), to execute tasks with (tool use) using your enterprise systems and data sources.
Amazon Bedrock is serverless, which means that you don’t have to manage any infrastructure. You can securely integrate generative AI capabilities into your applications using the AWS services that you are already familiar with.
Highlights
- Choose from a range of leading FMs for your use case
- Privately customize FMs with your own data using techniques like fine-tuning and Retrieval Augmented Generation (RAG)
- Build agents that execute complex tasks across your enterprise systems and data sources
- Integrate generative AI capabilities into your applications without managing infrastructure
About Amazon SageMaker
Amazon SageMaker is a fully managed service where you can build, train, and deploy ML models at scale using tools like notebooks, debuggers, profilers, pipelines, MLOps, and more — all in one integrated development environment (IDE).
While Bedrock is a platform focused on foundational models (FMs), SageMaker caters to a much broader range of machine learning (ML) models. Additionally, SageMaker offers greater control over the underlying infrastructure hosting the models.
Amazon SageMaker enables not only data scientists and developers, but also business and data analysts, regardless of their ML expertise, to build, train, and deploy machine learning models.
Highlights
- Choice of ML tools
- Fully managed, scalable infrastructure
- Repeatable and responsible ML workflows
- Human-in-th-loop capabilities
- Generative AI assistance with Amazon Q Developer (from data preparation and model training, to model deployment. code suggestions, troubleshooting assistance.
Leverage Lyzr Agents on AWS
With Lyzr on Amazon Bedrock you can deploy intelligent agents to perform tasks like competitor analysis, real-time data processing and customer interaction management.
The seamless integration with AWS services means you can get access to Lyzr’s AI capabilities without the complexity of deployment or infrastructure management.
Lyzr’s AI agents, including our autonomous agents for customer service, sales and operational workflows, are now supported on Amazon SageMaker.
You can deploy these agents on a range of hardware configurations to suit your use case. This supports fine tuning to help you customize Lyzr’s agents to your business needs.
Whether you need an AI agent for real-time customer support or advanced sales automation, deploying Lyzr on SageMaker means high performance, scalability and flexibility.
Lyzr’s Python Agent Framework for Multi-Agent Systems
The Lyzr Python Agent Framework makes deploying, configuring, and managing AI agents on AWS easy by integrating key components that streamline these processes.
It provides a single interface for developers and businesses to interact with Lyzr agents on both SageMaker and Bedrock.
With the Python framework, you can set up workflows that integrate Lyzr agents into your existing AWS environment so they work seamlessly across departments and functions.
Lyzr Agent API
Modular API framework for developing versatile LLM-based agents.
The Lyzr Agent API is a versatile and modular framework designed for developing role-based agents powered by Large Language Models (LLMs). It enables the creation of both chat applications and task automation systems, allowing agents to either engage in real-time conversations or execute complex, multi-step processes. With three core components—Environment, Agent, and Inference—developers can easily define, configure, and manage agents suited to various use cases. Whether building chatbots or automating tasks, Lyzr Agent API streamlines the development of intelligent, role-driven solutions for diverse applications. Read Documentation.
Now that you understand the what and why of Lyzr’s agent framework on AWS, let’s get started with Lyzr’s platform and products –
Lyzr’s Agentic Offerings on AWS
Enterprises can build Generative AI applications on Lyzr in any of the following ways based on their business requirements –
Pre-built Agents
Enterprises can build multi-task workflow automations with Lyzr’s pre-built AI agents.
- Jazon (The AI Sales Development Representative)
- Skott (The AI Marketer)
- Diane (The AI HR Expert)
- Kathy (The AI Competitor Analyst)
- Devi (The AI Data Analyzer).
Let’s take a closer look at Jazon to understand how it operates.
About Jazon: World’s 1st AI SDR
Jazon is like a super smart digital assistant that can help you promote and sell your products way more effectively than you could on your own.
Imagine you have a cool product that you want lots of people to buy. Normally, you’d have to spend tons of time researching potential customers, reaching out to them one-by-one, and following up over and over. It’s a super tedious process!
But with Jazon, all that hard work is automated. Jazon uses artificial intelligence to quickly find the right people who might want your product. Then it personally emails each of them with a customized message explaining why they’d love your product.
Jazon can handle doing this outreach for literally thousands of potential customers every single day – way more than any human could do. And it keeps following up automatically until people respond.
The best part is Jazon runs on your own computer systems, so your private data stays secure. It’s also designed to communicate in a friendly, empathetic way that people actually like.
So instead of wasting time on tedious tasks, you can focus on complex work that requires more of your attention and involvement while Jazon brings in more sales than you could ever get on your own.
Here are its features:
• Automates research, outreach, and responses
•Handles personalized outreach at massive scale (1000x more than humans)
• Runs locally on your cloud for 100% data privacy and security
• Inbuilt toxicity controller for empathetic, brand-aligned communication
• Completely customizable with your own data and prompts
• Continuously optimizes content, subject lines, send times etc.
• Automated follow-ups and nurturing sequences
• Detailed analytics on email performance metrics
• Integrates with existing marketing/sales tools
• Ensures email accessibility and deliverability
• Creates visually appealing, on-brand email designs
• Leverages AI for hyper-personalization at individual level
• Identifies best audience segments for targeted campaigns
• Streamlines entire email marketing workflow
How Jazon works
Jazon, the AI SDR, needs to initially be configured with information about your business, product, brand tonality, pain points, goals and other data you consider relevant for outreach. The next step is to upload the current lead list into Jazon, or connect Jazon to your CRM. You can either upload leads from the CRM or from a lead list, CSV file. Once that’s done, Jazon will go and start searching the internet and web scraping to find a lot more information about your leads. Information about their company, a website, perhaps even LinkedIn, anything Jazon can find about the lead.
It will then enrich that lead and gather all the information about the lead in Lyzr. Then Jazon begins writing personalized emails to begin the outreach. Those personalized emails include a few standard and specific details –
1. Information about your lead.
2. Your product information, your custom instructions, and examples of email structures or templates that have worked very well in the past.
Each time Jazon gets a positive reply afterwards or books a meeting, handles an objection, it feeds itself and learns over and over again, which means there’s no need for A-B testing because Jazon constantly iterates based on the feedback it gets. Like if it sends a type of email and gets higher reply rates, Jazon will send that email again and again, and then test.
If it gets lower reply rates, it will come back and it always iterates on its own results. So you don’t need to input any A-B tests or anything. It will personalize everything based on your leads data, product information, custom instructions, and examples it has from the past. Jazon also tests different personalization points, pain points, till it finds the best and then uses it for the best results.
Based on the email configuration settings, your infrastructure, Jazon figures how many emails to send. This is the warmup. And then you send the emails to your prospects.
If for some reason your prospects do not respond to any of the emails you send, if you set up 5 follow-ups, 20 follow-ups, which you shouldn’t do. If they don’t respond to any of those follow-ups, Jazon will then call them. They will call to check in with the lead, send them a voice message.
Now let’s say your prospect does reply, but has some objections, some questions about your product, then Jazon will dig into the FAQs provided about your product and answers to handle those objections, to answer the questions from your prospect so they would end up booking a sales call. It is a simple process of you getting the leads and Jazon doing everything, researching about them, crafting personal emails, sending those emails, handling the objections and replies, and booking in the meetings.
That’s what Jazon does. And all of that for $1,999.99. With a simple interface and all of it is happening on the cloud. You’re not sacrificing security.
Everything is happening on your cloud. No one is scraping and keeping your data about your leads, about whatever leads you scraped, about your result. Everything is happening in-house, so you’re not risking data security or privacy.
You’re not compromising anything, but you still get an AI SDR that rapidly books meetings by automatically handling objections, writing personalized emails based on your product settings, examples, and information about your customer.
Single Task Agents
Lyzr’s Single Task Agents are designed to automate specific, focused tasks within workflows. These agents handle a single operation at a time, ensuring precision and efficiency in areas like customer service automation, document review, or data analysis. For example, a single task agent can be deployed to automate answering customer queries by accessing knowledge bases or to handle SQL queries in response to natural language inputs.
Some popular use cases for single-task agents include:
| Use Case | Description |
| Chat Agent | State-of-the-art chatbots with short-term and long-term memory, RLHF (Reinforced Learning Human Feedback), a customizable system prompt, and seamless switches between LLMs. Try Lyzr Chat Agent |
| Knowledge Search | Bring perplexity-style document and data search to your organization. Lyzr’s document search can handle 100,000 documents and more, retrieving information accurately every time. And you get a citation to the source along with fetching the source file. Try Lyzr Knowledge Search |
| RAG Powered Apps | It is probably the most comprehensive RAG agent ever built. Just by tuning the parameters, you can choose up to 630 different RAG pipelines. You can also integrate with any LLM, vector store, embedding model, reranker, or parser of your choice. Read Blog |
| QA Bot | A simplified version of Lyzr Chatbot—an accurate question-answering agent to power your help and FAQs section. QA Bots are also the best fit for handling one-way conversations like employee handbook searches, law lookup searches, etc. |
| Data Analysis | Build a conversational analytics app on your data in a pythonic way with Lyzr’s Data Analyzr agent. The text query converts into a python code that runs on a pandas dataframe invoking several ML libraries like Scikit-Learn, StatsModel etc, helping you run various analysis on your data including regression, correlation, clustering, time series and more. |
| Text-to-SQL | Want to run an exploratory analysis on your database? Run Lyzr’s Text-to-SQL agent to convert your text queries to SQL code that can handle complex joins, extract information, and build insights and recommendations. |
These agents integrate seamlessly with enterprise applications, offering scalability and privacy by running locally on users’ cloud environments (AWS).
Agent API Studio
The Agent API Studio from Lyzr is a powerful, low-code platform that simplifies the development and deployment of AI agents. Launched at PyCon 2024, this tool allows both developers and non-technical users to build AI agents quickly, without requiring deep technical expertise. The platform is designed to support multiple industries, enabling tasks such as sales automation, customer service, and data analysis.
Highlights of Agent API Studio:
- Ease of use: It provides a straightforward interface for building and scaling AI agents with minimal code, making it accessible to a broad range of users.
- Multi-agent architecture: Supports the creation of multiple autonomous agents, which can handle complex workflows across different domains.
- Enterprise-grade tools: The platform includes advanced monitoring, logging, and real-time documentation, making it suitable for enterprise applications
The Agent API Studio was designed to encourage widespread AI adoption by making it easy to integrate into existing infrastructures, with built-in support for popular cloud platforms like AWS and Google Cloud. Read Documentation.
Agent API Studio Dashboard reflects user statistics and API details
In the Agent Builder start building –
- Create the Environment
- Configure the LLM
- Enable Modules (Short-Term Memory, Long-Term Memory, RAG KB/Chat Agent, Humanizer)
- Configure Tools (Perplexity Tool for Research, LinkedIn Tool for Posting, Mail Tool for Emailing)
- Configure the Agent
- Create Agent Interaction Sessions for the agent to draw Inferences
Setup the RAG with Vector store, API keys
Apps is the marketplace within the studio where all the published agents are displayed.
Custom Workflow
Lyzr’s solution for Curatal, a client in the business of talent acquisition for the IT industry, is a case in point of how custom workflows can be designed, developed on Lyzr’s agent framework and implemented on AWS based on specific client requirements.
Curatal provides the following services to businesses –
- Assessment Platform – A talent management and recruitment platform that tests coding skills.
- Comprehensive Testing Solutions – For large organizations like Infosys and Wipro
- Engaging experiences for Candidates – Mock tests, daily questions, and interactive activities
The company was facing challenges related to –
- Mounting Operational Costs due to increase in interviewer hires and inefficient, error-prone manual internal processes.
- Limited Scalability due to limited interviews conducted by interviewers available.
- Inconsistencies in Candidate Engagement
Lyzr’s Solution Offering to Curatal was a customized suite of agents deployed behind systems and modules.
Lyzr Agents Powering The Workflow
Curatal’s AI Stack – Lyzr x Amazon x OpenAI
Solution Implementation
Interview Management System
A robust platform that automates the scheduling, management, and execution of interviews. Supports various interview formats, including mock interviews, live interviews, and AI-assisted interviews.
Question Generation Module
Utilizes a combination of pre-existing question banks and real-time AI capabilities to generate interview questions dynamically. The system adapts to the candidate’s skill level, ensuring that questions remain challenging yet appropriate.
Answer Evaluation Module
A powerful tool for assessing candidate responses in real-time. This module evaluates both subjective answers and coding exercises. It includes a runtime environment for executing code, checking for errors, and providing detailed feedback on performance
Benefits
Lyzr’s solution led to enhanced outcomes for the client
1. Automated Interviews – Mock and live interviews, fully automated or AI-assisted
2. Adaptive Interviews – Dynamically adjust question difficulty based on responses
3. interviewer Aids – AI tools generate questions, evaluate responses, and offer suggestions
Impact of Lyzr AI
1. Lyzr AI has significantly improved Curatal’s recruitment and talent management processes.
2. Increased productivity, enhanced candidate engagement, improved efficiency and consistency.
3. Curatal has seen a substantial increase in the number of interviews conducted.
Client Testimonial
“Partnering with Lyzr AI has been a game-changer for Curatal. Their AI-driven solutions have transformed our recruitment process, enabling us to conduct more interviews, better engage with candidates, and streamline our operations. The adaptive interview capabilities and real-time evaluation have significantly improved our ability to assess talent efficiently. We look forward to further innovations from Lyzr to enhance our platform.”
— Curatal Team
Lyzr’s autonomous agents with AWS’s infrastructure gives businesses a way to streamline operations, make better decisions and improve customer experiences. By connecting AI agents across all functions Lyzr is helping businesses break down silos and get more efficient. Additionally, AI agents can help businesses analyze financial data to make better decisions and improve efficiency.
Interested? Contact Lyzr today to schedule a demo and see Lyzr and AWS in action.
Book A Demo: Click Here
Join our Slack: Click Here
Link to our GitHub: Click Here
