Customer service
AI triage for the support inbox
Classifies incoming emails and tickets, drafts answers from your knowledge base and routes edge cases to the right person with full context.
- Python
- LLM API
- Zendesk or Freshdesk
- RAG
Forward deployed engineering
Forward deployed engineers who build AI inside your team, not beside it.
encorp.ai engineers join your team, work in your repositories, tickets and cloud, write production code and own the result until it runs — then hand it over so your people can run it. Based in Sofia, working in Bulgarian and English within EU working hours, at a fraction of the cost of hiring a forward deployed engineer in the US.
650+
AI automations deployed by encorp.ai
2–4 wk
from kickoff to first results
EN/BG
engineers in Sofia, within EU working hours
Embedded FDE · Support triage
in production- 1Day 1Accounts in your GitLab, Jira and AWS; joins your team's daily stand-up
- 2Week 1Shadows the support team and agrees the outcome: AI triage for every inbound email
- 3Weeks 2–3Ships the triage service through your CI/CD, with evals and a human-review queue
- 4HandoverRunbook, dashboards and pairing sessions — your engineers now run the code
Your repositories · your cloud · your release process
We work in your stack, not ours
- GitHub & GitLab
- Azure DevOps
- Jira & Confluence
- Slack & Microsoft Teams
- AWS
- Microsoft Azure
- Google Cloud
- Kubernetes
- Your ERP & CRM
- On-premise systems
The short answer
What is a forward deployed engineer?
A forward deployed engineer (FDE) is a senior software engineer embedded in one customer's team who writes production code inside that customer's own systems and tools, connects it to their data and owns the outcome until it runs in production. The model comes from Palantir, where the official job title is Forward Deployed Software Engineer.
At Palantir, these engineers are known internally as “Deltas”: they deploy the Foundry and Gotham platforms at customers, while “Devs” build the platforms themselves. Palantir says a Delta does significantly more engineering than a consultant. The simplest way to place the role: the sales engineer helps sign the deal, the solutions engineer designs it, and the forward deployed engineer makes it work in production — then feeds what they learn back to the product.
encorp.ai offers the same engagement model to companies in Bulgaria and across the EU, without the cost of hiring a forward deployed engineer in the US. Our forward deployed AI engineers work from Sofia inside your repositories, ticket system and cloud, in Bulgarian or English, within EU working hours and aligned with the EU AI Act. Clients buy it as FDE as a service: a fixed-price pilot sprint first, then an embedded team for as long as the roadmap needs it.
Nikola Alexandrov, CEO of Hydra Blockchain, describes the engineers “we've onboarded through them” as “not only skilled but also proactive, seamlessly integrating into our projects.”
Why the model is booming
AI pilots are easy to start. Production is where they stall.
MIT NANDA's “The GenAI Divide: State of AI in Business 2025” found: “Despite $30–40 billion in enterprise investment into GenAI… 95% of organizations are getting zero return.” In our experience, what is missing is rarely the model but the engineering work that connects it to real systems, data and workflows. That is why the largest AI vendors now send engineers to their customers: AWS announced a $1B Forward Deployed Engineering unit on 30 June 2026, Microsoft launched Microsoft Frontier Company with $2.5B on 2 July 2026, OpenAI launched the OpenAI Deployment Company with $4B, led by TPG, on 11 May 2026, and Salesforce has committed to 1,000 forward deployed engineers.
~67%
deployment rate when companies partnered with external vendors, vs ~33% for internal builds (MIT NANDA, 2025)
5%
of organisations evaluating enterprise-grade GenAI tools reached production (MIT NANDA, 2025)
~729%
more forward deployed engineer job postings in April 2026 than a year earlier (Indeed data via Business Insider, 2026)
What embedded engineers do
Engineers who own the result, not just the recommendation
An FDE engagement combines three jobs that usually sit with different vendors: understanding the workflow, building the software and making sure your team can run it.
Workflow discovery with your people
We sit with the team that does the work, measure where the hours go and agree one outcome that matters to the business before writing any code.
Production code in your repositories
Code lives in your GitHub, GitLab or Azure DevOps, follows your review rules, ships through your CI/CD and deploys to your cloud or servers.
AI integration with the systems you have
APIs, databases, ERPs, CRMs and legacy applications connected to LLMs, retrieval and AI agents, with least-privilege access to each system.
Evals, security and the EU AI Act
Test sets and evals for AI quality, logging, human review where decisions matter, and documentation that supports your EU AI Act obligations.
Outcome ownership and weekly demos
One agreed outcome, a working demo on your data every week and a named owner on our side who answers for delivery.
Handover and enablement
Runbooks, architecture notes and pairing sessions with your engineers — plus optional AI-OPS monitoring if you want us to keep watching production.
What FDEs ship
What embedded engineers deliver in the first 30–90 days
Typical first deliverables of an embedded engagement. These are examples of the work, not client case studies — your scope is set after the fit call.
Finance
Invoice and document extraction into the ERP
Turns supplier invoices, bank statements and receipts into validated ERP entries, with an exceptions queue for anything the model is unsure about.
- Document AI
- ERP API
- PostgreSQL
- Review UI
Compliance
Onboarding and KYC document checks
Checks customer documents and forms against your onboarding checklist, flags missing or inconsistent data and keeps an audit log for compliance.
- Python
- Document AI
- Audit log
- Your KYC provider
Sales
CRM enrichment and follow-up agent
Researches new leads, fills missing CRM fields, summarises calls and drafts follow-ups that a salesperson approves before sending.
- HubSpot or Salesforce
- Node.js
- LLM API
- MCP
Operations
Internal knowledge assistant with permissions
Answers staff questions from Confluence, SharePoint and file shares while respecting the access rights each employee already has.
- RAG
- Vector database
- SharePoint / Confluence
- SSO
Engineering
Evaluation and monitoring harness
Test sets, automated evals, cost and latency dashboards and alerts, so AI features stay reliable after our engineers hand over.
- Evals
- CI/CD
- OpenTelemetry
- Grafana
Choose the delivery model
Consultancy vs hiring in-house vs forward deployed engineers
Three ways to get an AI initiative into production. Each fits a different situation; this is how they compare on the points that usually decide it.
| Consultancy | Hiring in-house | Forward deployed engineers (encorp.ai) | |
|---|---|---|---|
| Time to start | Weeks, usually after a paid discovery phase | Months to recruit, hire and onboard | Scoped on a free call; first results in 2–4 weeks |
| Writes production code | Rarely — delivers analysis and recommendations | Yes, once the team is hired | Yes — in your repositories from the first sprint |
| Owns the outcome | Owns the advice; delivery stays with you | Yes, alongside every other priority | Yes — an outcome agreed before work starts |
| Cost | Day rates, often with an open-ended scope | Salaries, recruitment and benefits (US FDE base at OpenAI: $185K–$325K) | Fixed-price pilot, then monthly — a fraction of a US hire |
| Knowledge transfer | Reports and presentations | Stays in-house, but leaves when people leave | Code, tests, runbooks and pairing with your team |
| Flexibility to scale down | Bound to the project contract | Low — employment contracts and notice periods | Team size reviewed and adjusted month by month |
How an engagement runs
From fit call to handover: how our forward deployed engineers work
Short cycles, your tools from day one, and a working release before anyone talks about scaling.
01
30 min
FDE fit call
We look at the workflow, your stack, your team and your security constraints, and tell you honestly whether embedded engineers are the right model.
- Fit or no-fit recommendation
- Candidate workflow and outcome
- Access and security checklist
02
Weeks 1–2
Embed and map
Engineers get accounts in your tools, join your stand-ups and shadow the people who do the work. Together we fix one outcome and how it will be measured.
- Agreed outcome metric
- Architecture note
- Backlog in your ticket system
03
Weeks 2–4
Build and ship
Production code in your repository, reviewed by your engineers, tested with evals on real (sanitised) data and released under supervision.
- Merged code in your repository
- Evaluation and test suite
- First release in production
04
Month 2+
Hand over or scale
We document, pair with your engineers and hand over — or continue as an embedded team on the next workflow, sized to your roadmap.
- Runbooks and documentation
- Pairing sessions with your team
- Plan for the next workflow
Ways to work with us
Start with one workflow, then keep the engineers you need
No public day rates and no open-ended retainers. The fit call is free, a pilot sprint gets a fixed scope and price, and an embedded team is scoped month by month.
All prices exclude VAT.
FDE fit call
Find out whether embedded engineers suit your AI project
Free
30 minutes · online
- Review of the workflow, stack and team
- Fit check: FDE engagement, a single automation or another route
- Access and security requirements
- Clear go / no-go recommendation
FDE pilot sprint
One workflow, from your backlog to production
Custom
fixed scope and price agreed after the call
- Embedded engineers in your tools and meetings
- Production code in your repository
- Evals, security review and supervised go-live
- Handover documentation and a demo for your team
Embedded FDE team
Engineers who stay while the roadmap grows
Custom
monthly · scoped per team
- Engineers dedicated to your AI roadmap
- Scale the team up or down month by month
- Optional AI-OPS monitoring in production
- Regular knowledge transfer to your engineers
Final scope and price follow the fit call. US salary ranges on this page are posted base salaries from company job boards as of September 2026 and exclude equity.
Track record
650+ AI automations deployed for companies like these
Selected case studies from our portfolio in fintech, telecom, logistics, media and e-commerce.
Econt
Implemented a custom AI agent for Econt, the leading logistics provider in Bulgaria, to automate customer support.
View case study
vivacom
Vivacom is one of the leading telecommunications companies in Bulgaria, offering a full range of services including mobile, fixed-line, internet, and digital television. As part of the United Group, Vivacom is recognized for its innovation, reliable network infrastructure, and customer-centric approach. The company serves millions of residential and business clients, delivering high-speed connectivity, advanced telecom solutions, and cutting-edge digital services across the country. With a strong focus on digital transformation, 5G technology, and smart business solutions, Vivacom continues to play a key role in shaping the future of Bulgaria’s telecom sector.
View case study
Forbes Bulgaria
Delivered AI-focused educational trainings for Forbes Bulgaria to empower business leaders with practical knowledge.
View case study
FAQ
Forward deployed engineers — common questions
What is the difference between a forward deployed engineer and a consultant?
A consultant advises and recommends; a forward deployed engineer builds. The FDE writes production code inside your systems, connects it to your data and stays accountable until it runs in production. Palantir, where the role comes from, says its forward deployed engineers do significantly more engineering than consultants. In practice, you get working software and a team that can run it, not a report to implement yourself.
How is a forward deployed engineer different from a solutions engineer or a sales engineer?
They work at different stages. A sales engineer runs demos and proofs of concept before the sale and does not write production code. A solutions engineer scopes and designs the solution around the sale and onboarding, and rarely ships to production. A forward deployed engineer is embedded after the decision and makes the solution work in production. Put simply: the sales engineer helps sign the deal, the solutions engineer designs it, the FDE makes it work.
How much does a forward deployed engineer cost?
Hiring one in the US is expensive: as of September 2026, OpenAI posts a base salary range of $185K–$325K for the role in San Francisco and Anthropic $280K–$320K in the US, plus equity. encorp.ai engineers from Sofia cost a fraction of that, with no recruitment fees: the fit call is free, a pilot sprint gets a fixed price after the call and an embedded team is priced monthly per scope, excluding VAT.
How is an encorp FDE engagement structured?
In three steps. A free 30-minute fit call checks whether embedded engineers suit your workflow, stack and team. An FDE pilot sprint follows with a fixed scope and price: our engineers join your team, agree one outcome, ship it to production and hand it over. If the pilot delivers, you continue with an embedded FDE team, scoped monthly, that takes on the next workflows and grows with your roadmap.
Do you work in our tools and environment?
Yes — that is the point of the model. Our engineers use accounts you provide in your code repositories, ticket system, chat and cloud, follow your code-review and release rules, and deploy to your infrastructure. The code, data and documentation stay in your environment. If your security policy requires VPN access or restricted permissions, we agree the setup before the sprint starts.
Why are AWS, Microsoft and OpenAI investing in forward deployed engineers?
Because most AI projects stall between pilot and production. MIT NANDA's 2025 research found that only 5% of organisations evaluating enterprise-grade GenAI tools reached production. The vendors' answer is to embed engineers: AWS announced a $1B Forward Deployed Engineering unit in June 2026, Microsoft launched Microsoft Frontier Company with $2.5B in July 2026, OpenAI launched the OpenAI Deployment Company with $4B in May 2026, and Salesforce committed to 1,000 FDEs.
What happens to the code and knowledge when the engagement ends?
They stay with your team. Everything is built in your repositories and documented as we go: architecture notes, runbooks, test and evaluation sets, and dashboards. Before handover, our engineers run pairing sessions so your people can change and operate the system themselves. If you prefer, we keep monitoring it in production through our AI-OPS service or come back for the next workflow.
Is a forward deployed engineer the same as an AI frontier engineer?
No, but the roles overlap. “Forward deployed” describes where an engineer works: inside the customer's team. “Frontier” describes what they work on: frontier AI models and agents. Many engineers are both. This page covers the embedded engagement model; the skills needed to build with frontier models are covered on our AI frontier engineer page.
Can companies outside Bulgaria hire forward deployed engineers from encorp.ai?
Yes. encorp.ai works with companies in Bulgaria and across the EU. Our engineers are based in Sofia and work in Bulgarian or English within EU working hours, so they join your team's daily meetings and reviews. Most of the work is done remotely inside your tools; on-site days for a kickoff workshop or the handover can be agreed in the scope of the sprint.
How do you handle security, data protection (GDPR) and the EU AI Act?
We work under your security and data protection rules, not ours. Engineers get least-privilege access to the systems a sprint needs, personal data stays in your environment under your GDPR processes, and you can revoke access at any time. Our delivery is aligned with the EU AI Act: we document the AI system's purpose, data and risks, add human review where decisions affect people and set up logging for your compliance team.
Free fit call
Tell us which AI workflow needs engineers inside your team
Send a short description of the workflow, your stack and your team. A senior engineer replies within one working day with a first view on fit and team setup.
My experience with Encorp has been excellent, especially in terms of the quality and dedication of the engineers that we’ve onboarded through them.
Explore more
More AI platforms and engineering services
Microsoft Copilot Automation
Copilot Studio agents, agent flows and Power Automate automations built inside your Microsoft 365 tenant.
Microsoft Copilot AutomationMicrosoft Foundry Development
Pro-code AI agents and applications on Microsoft Foundry (formerly Azure AI Foundry), in EU regions.
Microsoft Foundry DevelopmentMicrosoft AI Platform
Architecture and delivery across the Microsoft AI stack: Microsoft 365 Copilot, Copilot Studio, Foundry, Azure OpenAI and Fabric.
Microsoft AI PlatformGoogle Gemini Enterprise
Gemini Enterprise agents and automations connected to Google Workspace, Microsoft 365 and your business systems.
Google Gemini EnterpriseAI Frontier Engineers
Senior engineers who turn frontier AI models into agents and automations running in production.
AI Frontier EngineersStop collecting AI pilots. Put engineers inside the work.
Book a free 30-minute fit call. We will look at one workflow and tell you whether embedded engineers, a single automation or another approach gets it to production fastest.
650+ AI automations deployed · Engineers in Sofia · Bulgarian and English · EU AI Act-aligned · Prices exclude VAT