Sagitta

BACKGROUND

Please describe your company and position.

I am the Co-Founder & COO of nocnoc

Describe what your company does in a single sentence.

nocnoc is the largest cross-border online store in Latin America. We’re a one-stop solution helping global brands and retailers make their e-commerce feel right at home in the region. We do the heavy lifting and take care of cultural differences, logistical loops, and financial hurdles.

With our solution, as a global retailer, you can sell your products across 15 marketplaces in less than 48 hours without having to open an account on each marketplace. Instead, your products are featured in our megastores, already a favorite among Latin Americans. 

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Sagitta to accomplish?

• NocNoc partnered with Sagitta to develop an automated product categorization solution to streamline product publishing on a major Mexican marketplace.

SOLUTION

How did you find Sagitta?

Referral

Why did you select Sagitta over others?

• Close to my geographic location
• Great culture fit
• Referred to me
• Company values aligned

How many teammates from Sagitta were assigned to this project?

1 Employee

Describe the scope of work in detail. Please include a summary of key deliverables.

Overview
nocnoc partnered with Sagitta to develop an automated product categorization solution to streamline product publishing on a major Mexican marketplace. The solution integrates Machine Learning (ML), a predictive microservice, and a data engineering pipeline to classify products efficiently and ensure alignment with the marketplace’s taxonomy.

Scope of Work

Data Acquisition & Processing: Collected and cleaned product data via web scraping, mapping it to the Mexican marketplace’s taxonomy for compatibility.

Machine Learning Development: Built a KNN-based model using product embeddings for similarity-based classification and integrated an LLM for zero-shot classification to improve accuracy.

System Architecture & Deployment: Developed a microservice-based API, a batch ETL pipeline, and a tracking system to monitor outputs and improve categorization. The production version was deployed as a predictive microservice.

Roadmap for Future Enhancements: Delivered a plan to optimize taxonomy validation, automate dataset updates, and scale the system.
Key Deliverables
✔ Product Categorization Model – ML-based classification system.
 ✔ LLM-Assisted Classification – Zero-shot prediction integration.
 ✔ ETL Pipeline – Automated data ingestion and model updates.
 ✔ Microservice API – Scalable categorization service.
 ✔ Roadmap – Plan for long-term improvements.

Measurable Outcomes
Categorized thousands of products with high classification precision, reducing manual effort and streamlining the publishing process.
Delivered a scalable system architecture, allowing for future expansion and adaptation.
Project Management & Execution
Sagitta maintained clear communication, delivered milestones on time, and effectively adapted to evolving needs. Their team balanced technical complexity with usability, ensuring a smooth deployment.

What Stood Out About Sagitta
Strong AI integration, combining vector-based search and LLM-powered classification for a flexible and accurate solution.
Proactive problem-solving, refining the model iteratively while focusing on business goals.
Engineering expertise in ML, data pipelines, and microservices, ensuring a robust and maintainable system.
Areas for Improvement

No major areas for improvement. While our internal support could have been stronger due to competing priorities, Sagitta still delivered excellent results.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

Classification Precision: models achieved over 99% precision.

Automation Efficiency: Reduced manual effort in product categorization

Scalability & Performance: Designed a robust system architecture that allows for future expansion as the volume of categorized products grows.

Processing Throughput: The API-based categorization system enables real-time product classification.

Business Impact: Improved time-to-market for product listings on the Mexican marketplace.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

Sagitta demonstrated strong project management throughout the engagement. They maintained clear communication, delivered milestones on time, and effectively adapted to evolving needs. Their team balanced technical complexity with usability, ensuring a smooth deployment. Despite some internal challenges on our side due to competing priorities, they remained proactive and delivered excellent results.

What was your primary form of communication with Sagitta?

• Virtual Meeting
• Email or Messaging App

What did you find most impressive or unique about this company?

Their proactive problem-solving approach, refining the model iteratively while keeping business goals in focus, was also impressive. Additionally, their engineering expertise in ML, data pipelines, and microservices ensured a robust and maintainable system.

Are there any areas for improvement or something Sagitta could have done differently?

There were no major areas for improvement. While our internal support could have been stronger due to competing priorities, Sagitta still delivered excellent results. They effectively managed challenges and adapted to evolving needs, ensuring a smooth deployment.

BACKGROUND

Introduce your business and what you do there.:
We are a high-tech Startup created between Silicon Valley and Uruguay with the mission of building cutting-edge, massive-scale GenAI-first products designed to create positive impact. We are primarily building an AI assistant for real world tasks localized to LATAM. Im the Head of Engineering here, in charge of the devolpement team and driving results and new features to our users.

OPPORTUNITY / CHALLENGE

What challenge were you trying to address?: 

we partnered with Sagitta to enhance the platform, addressing backend, frontend, machine learning, and data acquisition needs.

SOLUTION

What was the scope of their involvement and team dynamic?:

The scope of work and deliverables included:

Backend Development:

- Deploying Machine Learning models to improve personalization and query-answering capabilities, including web scraping, API integration, and GenAI-based approaches.
- Expanding functionality by integrating additional eCommerce marketplace APIs, broadening its product search database.
- Enhancing backend microservices for scalability and performance, using Python.

Web Frontend Development:

- Redesigning the website’s architecture and implementing a fresh design using CSS & Tailwind.
- Building responsive, user-friendly interfaces with React and optimizing performance and SEO using Next.js.
- Enabling multi-language support with automatic language detection for inclusivity.
- Integrating Prismic CMS for easy blog updates and ensuring seamless mobile responsiveness.
- Adding direct app store links and QR code functionality to streamline user conversion.

Web Scraping:

- Developing scalable data scraping pipelines using Scrapy and Google Cloud Dataflow to collect large volumes of key data for feature development.
- Efficiently handling over a billion web pages to supply business intelligence that could not be accessed in real time.

What's the status of this engagement?: Some areas were completed and delivered successfully (web scraping), while others are still ongoing projects (like web and backend development).

RESULTS & FEEDBACK

How did your relationship with your partner evolve?:
Sagitta demonstrated exceptional project management across all areas:

- Deliverables were consistently on time, and the team was highly responsive to our evolving needs.
- Clear and transparent communication, along with adaptability to shifting requirements, ensured smooth collaboration.
- Their structured approach to managing large-scale scraping, backend development, and frontend projects aligned seamlessly with our workflows.

Several aspects of Sagitta’s work stood out:

- Technical Expertise: The team’s proficiency in React, Next.js, Tailwind, Python, Scrapy, and Machine Learning enabled high-quality solutions for complex challenges.
- Versatility and Scalability: They seamlessly handled large-scale distributed systems, complex Machine Learning deployments, responsive UI development, and high-volume web scraping.
- Fidelity to Design and Vision: Sagitta’s attention to detail ensured the website met design goals while maintaining consistent functionality across devices and languages.
- Team Integration: Their ability to integrate into our team allowed for efficient collaboration and rapid progress.

In what ways can they improve?:
While Sagitta delivered excellent results, one area for improvement could be expanding their on-demand resource availability. There were a few instances where scaling up the team to meet unexpected surges in workload would have been beneficial, especially during critical deployment phases. This would have allowed for even faster turnarounds during high-demand periods.
What advice do you have for clients with similar needs to yours?: 
100% recommended.
Sagitta’s contributions were invaluable, driving significant improvements across backend, frontend, and data capabilities while exemplifying professionalism, adaptability, and technical excellence.

UPDATED REVIEW

This review was published 11/17/2024. This is not the first update to this review

Co-Founder & Head of Engineering, AI Assistant CompanyInformation technology

Montevideo, Uruguay

11-50 Employees

Dec 2023 - Dec 2024

$50,000 to $199,999

Verified
Project summary
Sagitta was hired by an AI assistant company to scrape over a billion pages targeting Latin American data sources. They built scraping pipelines and used Python with Scrapy on Google Cloud Platform's Dataflow.

Feedback summary
Sagitta's infrastructure setup and approach to data collection were reliable and cost-effective. They met deadlines consistently, and communication was smooth and responsive. Their adaptability and structured approach to progress reporting were remarkable. Overall, the engagement was a success. This review is an update by the client company. The original content is located below the new review.

Quote
"What stood out most about Sagitta was their ability to build a robust, high-capacity scraping solution."

BACKGROUND

Please describe your company and position.

I am the Co-Founder & Head of Engineering of an AI assistant company. 

Describe what your company does in a single sentence.

We are building LATAMs most loved AI assistant.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Sagitta to accomplish?

• Scrape over a billion pages targeting latin american data sources

SOLUTION

How did you find Sagitta?

Referral

Why did you select Sagitta over others?

• Great culture fit
• Good value for cost
• Referred to me
• Company values aligned

How many teammates from Sagitta were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

The web scraping project with Sagitta was defined by a straightforward yet essential scope: to collect large volumes of key data needed to enhance our business and market intelligence. Sagitta’s role was to leverage Python with Scrapy on Google Cloud Platform's Dataflow to efficiently scrape over a billion web pages.

This data was critical for developing specific features, supplying necessary intelligence that could not be accessed in real time. Key deliverables included:

• Scalable data scraping pipelines tailored to handle high data volume.
• Collection of essential intelligence data to drive a specific feature, ensuring the system could access this information seamlessly when needed.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

One clear success metric from this project was our ability to efficiently scrape billions of data points at a very low cost.

Sagitta’s infrastructure setup and approach to data collection proved not only reliable but also cost-effective, allowing us to gather extensive intelligence data without inflating operational expenses.

This cost-efficiency has been crucial for sustaining and scaling our data capabilities within budget.

Web scraping at scale at low infrastructure cost

Describe their project management. Did they deliver items on time? How did they respond to your needs?

Sagitta displayed strong project management throughout this initiative.

They met deadlines consistently, and communication was smooth and responsive, especially when adjustments or clarifications were needed.

Their adaptability in scaling data scraping operations on demand, as well as their structured approach to progress reporting, aligned well with our requirements.

What was your primary form of communication with Sagitta?

• In-Person Meeting
• Virtual Meeting
• Email or Messaging App

What did you find most impressive or unique about this company?

What stood out most about Sagitta was their ability to build a robust, high-capacity scraping solution that integrated seamlessly with our existing infrastructure.

Their technical expertise and attention to detail in handling large-scale scraping projects were remarkable and contributed greatly to the success of this initiative.

Are there any areas for improvement or something Sagitta could have done differently?

An area where we see opportunity rather than a challenge is in the strategic prioritization of data sources. This collaboration highlighted the value of focusing efforts on the most impactful sources of data to maximize efficiency and relevance.

Moving forward, a joint effort between our team and Sagitta to refine the prioritization process could further enhance outcomes, ensuring we extract the most valuable intelligence while continuing to leverage Sagitta's expertise.
Overall rating: 5    

Quality: 5    

Cost: 5    

Schedule: 5    

Willing to refer: 5    

This review was published on 9/27/2024

Juan Olloniego
Co-Founder & Head of Engineering, BrainLogic AISoftware

Montevideo, Uruguay

11-50 Employees

Dec 2023 - Ongoing

$50,000 to $199,999

Verified

Project summary
Sagitta has helped a commerce AI company enhance their AI assistant capabilities. The team has deployed machine learning models, developed the solution’s backend and microservices, and integrated APIs.

Feedback summary
The client is pleased with Sagitta, which has successfully executed their tasks. The team has a strong project management approach, delivering tasks on time. Their members are responsive and adaptable to the client’s changing needs. Their workflow is transparent, resulting in a seamless engagement.

BACKGROUND

Please describe your company and position.

I am the Co-Founder & Head of Engineering of BrainLogic AI

Describe what your company does in a single sentence.

We are building Zapia an AI commerce assistant for LATAM, with over 2.5M users.

OPPORTUNITY / CHALLENGE

What specific goals or objectives did you hire Sagitta to accomplish?

• Enhancing AI assistant capabilities
• Re-doing the website from scratch
• Enhancing AI assistant capabilities

SOLUTION

How did you find Sagitta?

Referral

Why did you select Sagitta over others?

• Great culture fit
• Good value for cost
• Referred to me
• Company values aligned

How many teammates from Sagitta were assigned to this project?

2-5 Employees

Describe the scope of work in detail. Please include a summary of key deliverables.

Sagitta was engaged to augment our development team with their expertise in Machine Learning and Python backend development. Their primary focus was to enhance the capabilities of our AI-powered personal assistant, Zapia, which serves over 2.5M users in Latin America. Key deliverables included:

• Deploying Machine Learning models to personalize user experiences at scale.
• Integrating additional eCommerce marketplace APIs to broaden Zapia’s search scope.
• Improving Zapia’s ability to answer user queries through advanced information retrieval techniques, utilizing web scraping, API integration, and GenAI capabilities.
• Backend and microservices development in Python.

RESULTS & FEEDBACK

What were the measurable outcomes from the project that demonstrate progress or success?

The main outcome came due to increasing the development team size, speed of iteration and development of new functionalities. As mentioned earlier. Key deliverables included:

• The integration of additional API connections expanded Zapia's product search database, enhancing user satisfaction and search accuracy.
• The Machine Learning models improved the overall personalization and responsiveness of the assistant, contributing to increased engagement and retention rates.

Describe their project management. Did they deliver items on time? How did they respond to your needs?

Sagitta demonstrated strong project management skills throughout our collaboration. Deliverables were consistently provided on time, and they were responsive to our needs, even when timelines were tight. They adapted quickly to evolving project requirements, communicated effectively, and ensured transparency in their workflow.

What was your primary form of communication with Sagitta?

Email or Messaging App

What did you find most impressive or unique about this company?

What stood out the most about Sagitta was their technical depth and versatility. They not only provided solutions for immediate problems but also helped us refine our approach to scaling Zapia. Their ability to manage both large-scale distributed systems and complex Machine Learning deployments, all while understanding the nuances of Conversational Systems, was invaluable. 

Additionally, their seamless integration into our team made the collaboration feel natural and effective.

Are there any areas for improvement or something Sagitta could have done differently?

While Sagitta delivered excellent results, one area for improvement could be expanding their on-demand resource availability. There were a few instances where scaling up the team to meet unexpected surges in workload would have been beneficial, especially during critical deployment phases. This would have allowed for even faster turnarounds during high-demand periods.
Overall rating: 5    
They are very hard working, deliver results on time and can adapt easily to changing requirements.

Quality: 5    

Cost: 5    

Schedule: 5    

Willing to refer: 5