LLM-Powered Recommendation System

An Israeli startup is transforming U.S. service providers' personalized offerings. Dataforest scaled the project from prototype to a full web app with advanced ML, LLMs, and RAG fine-tuning. Managing 100,000+ products for 50,000+ customers, it delivers precise recommendations and revenue forecasts, maximizing sales opportunities

Marketing

Web and mobile development

Business process automation

Marketing

Web and mobile development

Business process automation

< 1 min

tailored recommendations delivery

100,000+

products supported by the platform

About the client

The client, an innovative Israeli startup, aims to transform how U.S. service providers—especially in insurance and financial services—generate business recommendations in complex environments.

Tech stack

Qdrant

Django

AWS

ReactJS

Claude 3 Sonnet

The client's needs

The client envisioned a recommendation system powered by LLMs to deliver personalized product suggestions based on customer demographics, needs, and global trends. They saw an opportunity to automate business offers in complex environments. Dataforest took this vision from prototype to a fully functional web app with sophisticated recommendation and forecasting tools.

‍

Challenges & solutions

Challenge

Develop a solution capable of handling an inventory scaled to over 100,000 products.

Solution

To address this complexity, we integrated the LLM with a Qdrant vector database and advanced ML models, including classification and clustering, further optimized results, providing precise recommendations at scale.

Challenge

Ensure the solution delivers results with high accuracy.

Solution

To deliver high accuracy for business offers generation, we integrated Qdrant vector DB with an LLM (Claude Sonnet 3) to perform retrieval-augmented generation (RAG), ensuring the system retrieves and contextualizes relevant data points before generating personalized product recommendations.

Engineered prompts tailored to yield highly accurate, contextually relevant recommendations and enhance the reasoning capabilities of the LLM, utilizing state of the art prompting techniques such as Chain-of-Thought (CoT) and Meta prompting.

Challenge

Achieve high efficiency and rapid results in generating new opportunities.

Solution

Our solution harnesses the power of Qdrant vector DB and advanced prompt engineering to deliver swift, precise filtering—driving exceptional speed and efficiency in opportunity generation.

Challenge

Delivering highly personalized, timely recommendations in an intuitive manner.

Solution

Developed a text2query system that transforms natural language product and customer descriptions into a custom ElasticSearch-like query format, enabling users to intuitively filter products and customer profiles for enhanced interaction flexibility.

Incorporated information on global events via Web Search services, allowing the recommendation engine to adjust suggestions based on relevant and impactful current events.

Leveraged the LLM to dynamically create and populate nodes within a knowledge graph, enabling a structured and interconnected representation of customer preferences, product information, and interactions, further enhancing recommendation quality.

Challenge

Solution

Challenge

Solution

Challenge

Solution

Challenge

Solution

Results

The client required a scalable recommendation system to suggest products based on customer characteristics and external trends, capable of handling from a few dozen to over 100,000 items.

Dataforest developed a robust solution integrating LLMs with a vector database, delivering personalized recommendations for up to 50,000 customers. Enhanced with advanced algorithms, this adaptable web app seamlessly fits various industries. Utilizing state-of-the-art prompting, it provides data-driven recommendations, accurate forecasts, and actionable insights—empowering businesses to offer precise and trend-focused solutions.

LLM-Powered Recommendation System

Steps of providing
data scraping services

Step 1 of 5

Free consultation

It's a good time to get info about each other, share values and discuss your project in detail. We will advise you on a solution and try to help to understand if we are a perfect match for you.

Step 2 of 5

Discovering and feasibility analysis

One of our core values is flexibility, hence we work with either one page high level requirements or with a full pack of tech docs. At this stage, we need to ensure that we understand the full scope of the project. Receive from you or perform a set of interviews and prepare the following documents: list of features with detailed description and acceptance criteria; list of fields that need to be scraped, solution architecture. Ultimately we make a project plan which we strictly follow. We are a result-oriented company, and that is one of our core values as well.

Step 3 of 5

Solution development

At this stage, we develop the scraping engine core logic. We run multiple tests to ensure that the solution is working properly. We map the fields and run the scraping. While scraping, we keep the full raw data so the final model can be enlarged easily. Ultimately we store data in any database and run quality assurance tests.

Step 4 of 5

Data delivery

After quality assurance tests are completed, we deliver data and solutions to the client. Though we have over 15 years of expertise in data engineering, we expect client’s participation in the project. While developing and crawling data, we provide midterm results so you can always see where we are and provide us with feedback. By the way, a high-level of communication is also our core value.

Step 5 of 5

Support and continuous improvement

We understand how crucial the solutions that we code for our clients are! Our goal is to build long-term relations, so we provide guarantees and support agreements. What is more, we are always happy to assist with further developments and statistics show that for us, 97% of our clients return to us with new projects.

How we provide data integration solutions

Step 1 of 5

Free consultation

It's a good time to get info about each other, share values and discuss your project in detail. We will advise you on a solution and try to help to understand if we are a perfect match for you.

Step 2 of 5

Discovering and feasibility analysis

One of our core values is flexibility, hence we work with either one page high level requirements or with a full pack of tech docs.

At this stage, we need to ensure that we understand the full scope of the project. We receive from you or perform a set of interviews and prepare the following documents: integration pipeline (which data we should get and where to upload), process logic (how system should work); use cases and acceptance criteria; solution architecture. Ultimately we make a project plan which we strictly follow.

Step 3 of 5

Solution development

At this stage, we build ETL pipelines and necessary APIs to automate the process. We attract our DevOps team to build the most efficient and scalable solution. Ending up with unit tests and quality assurance tests to ensure that the solution is working properly. Focus on Results is one of our core values as well.

Step 4 of 5

Solution delivery

After quality assurance tests are completed, we deliver solutions to the client. Though we have over 15 years of expertise in data engineering, we are expecting client’s participation in the project. While developing the integration system, we provide midterm results so you can always see where we are and provide us with feedback. By the way, a high-level of communication is also our core value.

Step 5 of 5

Support and continuous improvement

Steps of providing web applications services

Step 1 of 7

Web development discovery

In the initial stage of the web-based development project, professional business analysts make detailed documentation of the project requirements and the approximate structure of the future web application. DATAFOREST is a custom web application development agency, guided by extensive experience in multiple industries. We give you detailed project documentation and then assemble the team according to your time and budget.

Step 2 of 7

UX and UI design

Based on your wishes, the needs of your target audience, and the best web application design and development practices, our UX and UI experts create an aesthetically pleasing and user-friendly interface for your app to satisfy even the most demanding users.

Step 3 of 7

Web-based application development

At DATAFOREST we are following the best programming design principles and approaches. Being a data engineering company, we build high load platforms, with a significant level of flexibility and result orientation. We keep our deadlines and follow SOC 2 compliance requirements.

Step 4 of 7

Integration

With DATAFOREST, integrating the application into your business won’t stop your processes for a minute. We provide seamless integration with your software infrastructure and ensure smooth operation in no time.

Step 5 of 7

Quality assurance

We use a multi-level quality assurance system to avoid any unforeseen issues. Working with DATAFOREST, you can be confident that your web app development service solutions arrive to the market polished and in full compliance with all certification requirements.

Step 6 of 7

24/7 support

Once a product is released to the market, it’s crucial to keep it running smoothly. That’s why our experts provide several models of post-release support to ensure application uptime and stable workflows, increasing user satisfaction.

Step 7 of 7

Web app continuous improvement

Every truly high-quality software product has to constantly evolve to keep up with the times. We understand this, and therefore we provide services for updating and refining our software, as well as introducing new features to meet the growing needs of your business and target audience.

The way we deal with your task and help achieve results

Step 1 of 5

Free consultation

It's a good time to get info about each other, share values and discuss your project in detail. We will advise you on a solution and try to help to understand if we are a perfect match for you.

Step 2 of 5

Discovering and feasibility analysis

One of our core values is flexibility, hence we work with either one page high level requirements or with a full pack of tech docs.

In Data Science, there are numerous models and approaches, so at this stage we perform a set of interviews in order to define project objectives. We elaborate and discuss a set of hypotheses and assumptions. We create solution architecture, a project plan, and a list of insights or features that we have to achieve.

Step 3 of 5

Solution development

The work starts with data gathering, data cleaning and analysis. Feature engineering helps to determine your target variable and build several models for the initial review. Further modeling requires validating results and selecting models for the further development. Ultimately, we interpret the results. Nevertheless, data modeling is about a process that requires lots of back and forth iterations. We are result focused, as it’s one of our core values as well.

Step 4 of 5

Solution delivery

Data Science solutions can be a list of insights or a variety of different models that consume data and return results. Though we have over 15 years of expertise in data engineering, we expect client’s participation in the project. While modeling, we provide midterm results so you can always see where we are and provide us with feedback. By the way, a high-level of communication is also our core value.

Step 5 of 5

Support and continuous improvement

The way we deal with your issue and achieve result

Free consultation

Step 1 of 5

It's a good time to get info about each other, share values and discuss your project in detail. We will advise you on a solution and try to help to understand if we are a perfect match for you.

Step 2 of 5

Discovering and feasibility analysis

One of our core values is flexibility, hence we work with either one page high level requirements or with a full pack of tech docs.

Depending on project objectives, DevOps activity requires auditing the current approach, running metrics measurement, performing monitoring and checking logs. By having a set of interviews, we ensure that we understand the full scope of the project. Ultimately we make a project plan which we strictly follow. We are a result-oriented DevOps service provider company, and that is one of our core values as well.

Step 3 of 5

Solution development

At this stage, our certified DevOps engineers refine the product backlog. We deliver great results within digital transformation, cost optimization, CI/CD setup, containerization, and, last but not least, monitoring and logging. We are a result focused company – it’s one of our core values.

Step 4 of 5

Solution delivery

After quality assurance tests are completed, we deliver solutions to the client. Though we have over 15 years of expertise in data engineering, we expect client’s participation in the project. By the way, a high-level of communication is also our core value.

Step 5 of 5

Support and continuous improvement

Success stories

Check out a few case studies that show why DATAFOREST will meet your business needs.

All Success Stories

Financial Intermediation Platform

The project aims to develop a deal origination platform for private equity investments in infrastructure-related sectors and involves building a secure, interactive B2B platform from scratch, with sign-up functionality to connect investment firms to proprietary investment opportunities. DATAFOREST built a highly-loaded platform and applied AI functionality to empower the application's development.

98%

model accuracy

100%

timely development

Enrico Cattabiani

Founder & CEO IDN, Infrastructure Deals Network

View case study

Financial Intermediation Platform preview

They understood our requirements, translated into actions rapidly, and adapted to requests easily.

Bank Data Analytics Platform

The Bank Data Analytics Platform project aims to develop a web-based application for the Client and its customers to query and analyze data relating to various banks and other financial institutions. The project involves building an interactive B2B web app with custom dashboards and analytics features, as well as using AI functionality to empower the application's development and analytics capabilities.

30+

system integrations

19%

CX boost

Stuart Theobald

Chairman Intellidex, Financial Services and Consulting company

View case study

They've had a quick grasp of what we are trying to do and delivered to our spec without a fuss.

Performance Optimization & Bottlenecks Elimination

The financial services company faced performance issues and bottlenecks on its financial platform and was looking for a DevOps partner to increase application performance, stability, and resilience, and reduce operational costs. DATAFOREST performed a technical audit of the current AWS infrastructure, created a bottleneck monitoring system, re-developed inefficient SQL queries and data pipelines, and implemented horizontal scaling and a microservice approach using Docker and Kubernetes.

1000%

performance boost

20%

cost optimization

Daniel Garner

CTO Flexium, FinTech company

View case study

Performance Optimization & Bottlenecks Elimination preview

The team of DATAFOREST is very skilled and equipped with high knowledge.

Chargeback Management B2B SaaS Platform

Chargeback is a powerful B2B SaaS platform designed to revolutionize chargeback management. With enterprise-grade API integrations (Ethoca, Verifi), seamless payment processor connections (Stripe, Shopify, Braintree and others), and real-time dispute alerts, it empowers merchants to stay ahead of chargebacks before they happen. Featuring dynamic billing and intuitive dashboards, Chargeback helps businesses retain more revenue, reduce operational costs, and enhance trust with payment processors and customers.

90%

of transactions related to the incoming alerts are successfully identified

75%

of identified transactions are automatically refunded

View case study