Reliable AI starts with reliable data
AI systems rarely fail because of the model. They fail because of uncertain data. Orboroi exists to take that uncertainty out of your training data.
Our story
Orboroi was founded in 2021 by a team with more than ten years of experience running outsourcing and data operations.
We started with specialised annotation for medical imaging and computer vision, added data validation and quality audits, and now support AI teams from first labels to production. Our clients work with a Swedish company under EU contracts, while the people labelling their data are our own trained employees in Dhaka.
The people accountable for your data
Leadership
Bimal Dey
FounderLeads Orboroi’s strategy and client partnerships, and is accountable for every engagement.
LinkedIn profile
Zahidul Islam
Co-FounderBuilds and runs the delivery operation: the teams, the processes and the quality standards behind every batch.
LinkedIn profile
Shafizur Rahman Seeam
DirectorOversees client projects and delivery, from the first guidelines to the final batch.
LinkedIn profileTwo locations.
One team.
Contracts, guidelines and your project lead in Stockholm. Annotation and quality review by our own full-time team in Dhaka.
Stockholm
Kista, Sweden · CET (UTC+1)
- Client management and your project lead
- Guideline design
- Contracts, NDAs and data processing agreements
- Guidelines
- Annotation
- Two-layer QA
- Delivery
Dhaka
Bangladesh · BST (UTC+6)
- Full-time annotators and QA reviewers
- Trained per domain, from medical imaging to robotics
- Access-controlled office, no phones at workstations
Working hours in Central European Time (CET)
Your morning overlaps with the Dhaka team's working day, so feedback reaches the annotators the same day.
What we stand for
- QualityA 99% first-batch acceptance rate, measured and reported with every delivery.
- TransparencyYou always know the status, quality and people behind your data.
- ScalabilityFrom pilot projects to dedicated teams, without a drop in quality.
- Data certaintyThe confidence to deploy models that hold up in production.
Work with us
We agree the batch size on a short call. If the quality isn't there, you've lost nothing.