D

deepset

Provides NLP AI tooling for enterprises

Forward Deployed Engineering Manager

Full-TimePosted on 10/22/2026
No salary listed
Mid
Berlin, Germany
RemoteRemote role; Berlin HQ is noted, with in-person team meetups.
Engineering ManagementEngineering Leadership

About the job

Requirements
  • Experience leading customer-facing engineering, professional services, or delivery teams in enterprise software, AI, or infrastructure.
  • Strong coaching instincts and the ability to develop engineers and build high-trust teams.
  • Technical depth in retrieval-augmented generation (retrieval, embeddings, re-ranking, and evaluation), cloud and self-hosted deployments, model optimization, application programming interfaces, integrations, and identity and access management/security fundamentals.
  • Commercial awareness of how scope, staffing, and execution affect utilization, revenue, and margin.
  • Clear written and verbal communication with technical and executive audiences, and comfort with ambiguity in a remote-first setup.
Responsibilities
  • Own team health and culture across the Advisory, Builder, Operating, and Support Forward Deployed Engineering profiles, creating a safe space for hard questions and enabling engineers to share patterns instead of working in silos.
  • Run 1:1s and career conversations so every Forward Deployed Engineer understands their standing and development opportunities.
  • Build the hiring and onboarding pipeline for a remote-first, asynchronous team.
  • Clear blockers quickly and escalate unresolved blockers before they become delivery risks.
  • Assign the appropriate Forward Deployed Engineering profile to each engagement and deliver against team utilization targets while protecting against burnout and idle capacity.
  • Review architecture and solution decisions across retrieval-augmented generation pipelines, retrieval quality, deployment topology, and production operations to maintain the technical bar.
  • Deliver engagements within scope and plan, follow delivery processes, and own resolution of customer escalations.
  • Partner with the Delivery Lead and Value Engineers on scope, feasibility, and staffing, with clear ownership at each handoff.
  • Feed field insights to Product and Engineering and build reusable patterns that reduce custom work.
Desired Qualifications
  • Experience running services or delivery within a SaaS or platform company transitioning to enterprise and self-hosted deployments.
  • Experience with OKR frameworks and squad-based delivery organizations.
  • Familiarity with agent-based systems, large language model platforms, and AI-native engineering tools.

About the company

Deepset builds developer tools to help create production-ready AI and NLP systems. Its Haystack open-source framework lets engineers design pipelines for tasks like semantic search, document processing, and retrieval-augmented generation, and it works with multiple models and providers. On top of Haystack, Deepset offers Haystack Enterprise Platform as a SaaS for prototyping, deployment, monitoring, and governance, with options for cloud, on-premise, or air-gapped setups. The company aims to help organizations build scalable AI-powered applications that can search and reason over large data sets.

Company Size

51-200

Company Stage

Early VC

Total Funding

$45.6M

Headquarters

Berlin, Germany

Founded

2018

Get referred to deepset

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • June 2026 HPE Unleash AI and Celonis partnerships expand regulated-sector distribution.
  • July 2026 NTT DATA and European Commission deployments validate deepset in government workflows.
  • September 2026 Materna collaboration and August 2026 compliance platform launch strengthen Germany-first demand.

What critics are saying

  • Open-source Haystack commoditizes deepset's moat if hyperscalers bundle equivalent agent tooling.
  • Enterprise AI budgets consolidate around OpenAI, Anthropic, Microsoft, and Celonis, squeezing standalone platforms.
  • If sovereign-AI deals stall after 2026 pilots, deepset's enterprise revenue stays procurement-dependent.

What makes deepset unique

  • Haystack stays vendor-agnostic, open-source, and production-focused for regulated enterprise AI.
  • September 2026: Haystack Enterprise Platform unifies prototyping, deployment, monitoring, and governance.
  • deepset targets sovereign, air-gapped deployments through HPE, Celonis, Mozilla, and NTT DATA partnerships.

Help us improve and share your feedback! Did you find this helpful?

Benefits

  • Remote Work Options
  • Flexible Work Hours
  • Paid Vacation
  • Paid Sick Leave
  • Stock Options
  • Professional Development Budget
  • Mental Health Support
  • Company Social Events

Growth & Insights and Company News

6 month growth

2%

1 year growth

3%

2 year growth

3%
PR Newswire
Oct 21st, 2026
deepset launches AI agent cost optimisation and custom models for government and defence

Berlin-based deepset, the company behind the open source Haystack framework, has announced new tools and services to cut the cost of running sovereign AI agents. It unveiled them on October 21, 2026, at NVIDIA GTC Berlin. Haystack Enterprise Platform now shows AI spend by model, agent, service, and team. Teams can compare models and configurations to see whether a change lowers cost while meeting quality needs. deepset also offers custom model services for government, defense, energy, and regulated industries. Its engineers select, adapt, or post-train smaller specialised models for repeated tasks such as classification and routing. Customers own the models built for them. "The goal is to complete the task at the lowest cost that meets your quality and security requirements," said chief executive officer Milos Rusic.

PR Newswire
Oct 21st, 2026
deepset unveils open architecture for self-improving, sovereign AI agents on NVIDIA

Berlin-based deepset, the company behind the open source Haystack AI framework, has unveiled an open reference architecture for self-improving AI agents. Built with NVIDIA open models and software, it runs in locally controlled and air-gapped environments, so sensitive data stays under an organisation's control. Two Haystack agents form a feedback loop. One performs the task; the other analyses the results and proposes changes. Each change is tested against set objectives, with approval, audit and rollback controls keeping people in charge of what reaches production. The first use case is cybersecurity. At NVIDIA GTC Berlin on October 22, deepset will demonstrate an agent that checks confidential source code for vulnerabilities and learns from earlier runs. Example code is available on deepset's GitHub page.

Investors Hangout
Jun 18th, 2026
deepset and HPE team up to boost sovereign AI deployments.

deepset and HPE team up to boost sovereign AI deployments. The marriage of governance and AI. From Silicon Valley to sovereign shores, AI is making waves, but the path ain't always smooth. Today, deepset and HPE are shaking things up a notch, building a robust bridge between AI's potential and the stringent demands of sovereign environments. If you're snoozing on this, you're missing out on the seismic shift in AI deployment strategies. AI deployment: the new battleground. The partnership isn't just another techie pow-wow - it's a tactical move in the evolving game of AI governance. With the Haystack Enterprise Platform backing deepset's play, organizations can now blitz through AI use case rollouts while keeping risk, regulations, and rights in check. AI and governance coming together like this? It's the future, folks. Investors Hangout, LLC is talking about deploying AI systems that get self-hosted safely within sovereign stickers and air-gapped barricades. All while keeping those sensitive data sets and models securely under lock and key. "With agentic AI moving into operational deployment, organizations need infrastructure and AI platforms they can control and trust," Milos Rusic, deepset's CEO, emphasized the need for sovereignty in today's AI stageplay. Unpacking the HPE connection. Let's not forget HPE's role in this whole dance. The HPE Unleash AI partner program, all shimmer and shine, paints a picture of engineered AI success that is as flexibly governed as it is robust. Yeah, Investors Hangout, LLC is talking about stacking up HPE Private Cloud AI with NVIDIA's wizardry for a system ready to nun-chuck its way into production AI. Why sovereign AI matters. In a world that can't seem to stop talking data breaches and cyber threats, having control over AI deployment is more important than ever. And if you're the betting type, wagering on deepset and HPE's capabilities to bolster that control might just see your portfolio climb. They're arming enterprises with the tools to enforce governance, auditability, and lifecycle management. This isn't just for the tech geeks or defense wonks. Sovereign AI extends its appeal to organizations ranging from the public sector to the boardroom table. Whether it's using AI for cybersecurity in classified operations or supporting mission-critical decision-making, this collaboration covers the gamut. There's More in the AI Future. deepset's pedigree isn't something to overlook. With prior collaborations involving heavyweights like the European Commission and the German Ministry of Research, Technology, and Space, they're no strangers to tailoring AI operations under strict governance. In essence, by navigating the rough AI waters hand in hand with HPE, deepset's setting a precedent. They're not only prepping enterprises for the present but future-proofing them for what's bound to be a tech-dominated era. "Together, HPE and deepset are delivering secure, governed AI solutions that combine enterprise-grade infrastructure with flexible and governed AI agent capabilities," chimed in Robin Braun of HPE, spelling out the essence of their collaboration. Closing thoughts: from aspiration to action. With the world getting tighter around data control, the deepset-HPE collaboration is a powerful move. But let's be clear - this partnership isn't an overnight fix. It's an evolving strategy where patience meets preparedness, and that blend might just be what's needed in today's high-stakes AI landscape. If you're running an enterprise, it's time to strap in and watch as these AI forces give the game a fresh spin.

WirtschaftsWoche
Jun 16th, 2026
Cooperation is Europe's answer to Palantir.

Cooperation is Europe's answer to Palantir. Celonis and Deepset are joining forces for a sovereign AI platform for government agencies and critical infrastructures. The goal is a market that has so far been dominated by the controversial US corporation. 16.06.2026 - 4:11 pm Celonis co-CEO Bastian Nominacher (center) in conversation at the digital conference TECH on June 1, 2026 in Heilbronn Photo: Marc-Steffen Unger

EE News Europe
Jun 16th, 2026
Celonis and deepset launch sovereign AI platform for critical operations.

Celonis and deepset launch sovereign AI platform for critical operations. News | June 16, 2026 By Asma Adhimi German software companies Celonis and deepset have teamed up to develop a sovereign AI platform aimed at mission-critical environments, including public sector organizations, defense, cybersecurity, policing, and critical infrastructure. The partnership combines process intelligence with AI agent orchestration to help organizations deploy AI systems while maintaining control over data, models, and infrastructure. For eeNews Europe readers, the announcement highlights the growing push for European-controlled AI technologies at a time when governments and critical infrastructure operators are seeking alternatives to non-European platforms. It also reflects increasing demand for AI systems that can operate in highly regulated and security-sensitive environments. European alternative for sovereign AI. The partnership brings together Celonis' Process Intelligence Platform and its recently introduced Celonis Context Model (CCM) with deepset's Haystack-based AI agent platform. The companies say the integrated platform is designed to provide organizations with a trusted AI environment that can unify operational data, documents, and workflows. The goal is to improve situational awareness, accelerate investigations, identify operational risks earlier, and support decision-making across departments and agencies. A key feature is deployment flexibility. According to the companies, the platform can run on an organization's preferred infrastructure without creating vendor lock-in, allowing users to retain full control over their data and AI operations. "At deepset, we believe organizations need AI systems they can fully trust and control - especially in mission-critical environments," said Milos Rusic, CEO and co-founder of deepset. "By combining Haystack's open and governed AI agents with the Celonis Context Model, we are creating a European alternative for operational AI that is transparent, auditable, and grounded in real-world operational context." Process intelligence meets AI orchestration. The collaboration addresses one of the major challenges facing enterprise AI deployments: providing systems with accurate operational context. Celonis says its Process Intelligence Platform and CCM allow AI systems to understand how organizations function in real time by connecting process data, business knowledge, and decision intelligence. The company argues that this operational layer is often missing from enterprise AI implementations. "AI systems are only as effective as the operational context they are given," said Florian Schewior, Managing Director DACH at Celonis. "By combining Celonis Process Intelligence and Context Model capabilities with deepset's AI orchestration platform, we can help public sector and security organizations deploy AI systems that are not only powerful, but operationally grounded, explainable, and aligned with sovereign infrastructure requirements." Focus on Europe's security and public sector markets. The initial focus of the partnership will be on organizations in the DACH region, while also supporting broader European sovereign AI initiatives linked to security, defense, and public sector modernization programs. The companies said they plan to expand their collaboration through joint customer projects, ecosystem partnerships, and industry events as demand for sovereign AI solutions continues to grow across Europe. Celonis, headquartered in Munich and New York, has built its reputation around process intelligence software, while Berlin-based deepset develops the open-source Haystack framework for building AI applications and agents. Together, they are positioning the new platform as a European-built alternative for organizations that require both advanced AI capabilities and strict control over operational data. Linked Articles