Forward-deployed AI engineering
Software that fits exactly one business → yours
We execute your AI transformation. Our tech team works from your office, creating apps on top of the data you already have.
The process
You know which process it is.
Not a category, a process. If one of these is yours, you already know roughly what it costs you and you have already tried to fix it twice.
The quote that takes half a day.
Out in minutes, priced the same way every time.
The report someone rebuilds by hand every week.
Built on a schedule, checked, and simply reviewed.
The exception that turns into a ticket.
Handled where it happens, with a person approving it.
The person who keeps it all in their head.
In a system your team can ask, not in someone's memory.
Whichever one it is, it goes the same way from here.
A diagnostic first, then a build around one real workflow, then your team running it. Pick a stage to see what happens in it.
We spend two to four weeks inside your operation, mapping the workflows, the data and where the time actually goes. It ends with the three to five places AI would earn the most in your business, sequenced, with the case for each one.
Talk to us
Start with the process that costs you most.
- 01You tell us which process is slowing the business down.
- 02We come back with how we would approach it, and whether we are the right people at all.
- 03If it fits, we scope the diagnostic together.

SOC 2 Type II · ISO 27001 · HIPAA · GDPR
How we work
We already run an AI-native business.
Not a pitch we assembled for this page. These are the products this team has built and shipped, and the systems we run our own company on. When one of them is close to what your operation needs, you get it built around your workflow rather than from scratch.
- 70%
- of our own work runs without a person driving it
- 1M+
- pieces of work our agents run every month
- 70.4
- tasks delivered per week, up from 25.3
- 2,000+
- AI engineers in the community we can staff from
- 1.6M+
- professionals taught to work with AI
SOC 2 Type II · ISO 27001 · HIPAA · GDPR
Part of Juma Labs. Backed by True Ventures.
What it is built on
We've solved the hard problems in production AI.
Proprietary tech and state-of-the-art know-how, tested at scale.
- 30M
- chats
- 1M+
- agents
- 100K
- users
General Purpose Computer Agents
Production-ready, tested with millions of tasks. Agents solve arbitrary tasks on virtual machines: planning, execution, long-running, persistence, self-healing.
Cost & Speed
44x lower100x lower cost and 10x faster creation of deliverables: PDFs, PowerPoint, Excel, HTML. Production-grade output at a fraction of traditional AI costs.
MCP Integration Abstraction
One-shots new connectors and wraps any API in a highly performant integration. Build once, connect everything, with no custom integration work per client.
Context Handling & Prompt Hierarchies
Multi-layer systems with tools, caching, skills, and compression that produce fast, efficient, and consistent results across complex workflows.
Multi-Modal Orchestration & Failover
Routing and reliability engineered across every major vendor. Zero-data-retention agreements. AWS, Azure, GCP. Open-source model deployment for specific tasks.
Analytical Pipelines
Pipelines that analyse massive data at scale with gen AI models, and improve over time. Self-optimising, not static.
Private VPC Deployments
Completely separated and autonomous agent architectures on private virtual clouds. Multiple live deployments, fully air-gapped where required.
AI Monetisation & Metering
Immutable credit ledger, per-model, per-tool and per-task billing configurations. Cost attribution and spend observability built in from day one.
AI QA, Evals & LLM Observability
Full quality assurance and evaluation frameworks. LLM observability and RAG infrastructure, so every output is traceable, measurable, and improvable.
Who you would work with
The people you would be building with.
Juma Ops is the forward-deployed engineering team of Juma Labs, and this is the lab. The same people who build our own products embed with your operation: the founders lead the engagement, and the whole product-engineering team is here, not a highlights reel.

Iliya Valchanov
CEO

Ilko Kacharov
CTO

Yavor Belakov
Co-founder

Katya Vaptsarova
Co-founder & CFO

Simona Dobreva
Head of Product

Mario Stoev
Product
Show the rest of the teamHide the rest of the team

Antonio Georgiev
AI Engineer

Biser Atanasov
Senior Full-Stack Engineer

Ivo Kacharov
Senior Full-Stack Engineer

Ventsislav Staykov
Senior Full-Stack Engineer
Atanas Pepechkov
Head of Infrastructure
Kremena Shtereva
Senior QA Automation Engineer

Maria Valchanova
Co-founder, Brand & Design

Margarita Arsova
Founding GTM

Nadya Tabutova
People & Operations

Ivan Vaptsarov
Project & Operations Management Consultant