Building the default operating system for AI-driven enterprises.

We believe the next generation of companies won't hire large commercial teams. They'll deploy AI agents that plan, execute, and learn, operating sales, marketing, and customer service autonomously, around the clock.

Vision

The Problem

Revenue growth requires headcount growth. Every new market, campaign, or product launch means hiring more people. More admin, more coordination, more overhead. Tools don't solve this. They add complexity while the fundamental constraint remains: people are expensive, slow to onboard, and impossible to scale linearly.

Our Answer

AI employees. Not chatbots, not copilots, not dashboards with AI features. Autonomous agents that own entire workflows: prospecting, outreach, pipeline management, customer service. They operate continuously with human oversight. One platform that replaces the need for large sales, marketing, and support teams.

Intelligence

The world's commercial knowledge. And yours.

Two knowledge domains, unified. The world's largest commercial intelligence layer cross-referenced with deep knowledge of your own organization. Growing every day.

7M+

Companies

and growing daily

14M+

Contacts

linked to companies

100+

Fields

per company record

<100ms

Search

full-text, any query

Index

The world outside

7 million company records with org charts, technology stacks, hiring signals, funding, revenue, social profiles. 14 million contacts with job titles, emails, professional profiles. Full-text search across 14 million documents in milliseconds. Lead scoring against your ideal customer profile. Website scraping for real-time research. Data quality scoring on a 0-6 scale.

Yeda

Your world inside

The self-building knowledge base. Salesteq automatically researches your company identity, maps your products, builds your ICP, extracts knowledge from uploaded documents, reads your emails, knows your calendar. PDF extraction, conversational learning, six content categories. The AI gets smarter about your business with every interaction.

Kol

The voice

Proprietary speech engine built in-house by our co-founder. End-to-end pipeline: recognition, neural voice synthesis, real-time sentiment analysis, speaker diarization. Natural phone and chat conversations between AI agents and your customers. Every language. Every accent. Sub-second latency.

Zera

Zero to operational

The autonomous deployment engine. Salesteq lands in a new company like a probe on a foreign planet. It scans the company website, ingests product catalogs, imports existing contacts and deals, learns the ICP from historical data, and configures itself for operations. No integrations. No setup project. No consultants. From zero to operational in hours, not months.

The Combination

External world knowledge plus internal organizational knowledge equals total intelligence. No other platform has both. Salesteq knows everything about every company in the world AND everything about your company, your products, your pipeline, your conversations. This is what enables truly autonomous operations.

Badger

Relentless. Autonomous. Unstoppable.

Named after the Honey Badger. The autonomous agent at the core of Salesteq. Users interact through natural language. Badger plans, reasons, calls tools, spawns sub-agents, and executes. Not a chatbot. Not a copilot. An AI employee that operates your entire commercial stack.

01

45+ AI Tools

Intelligence: search 7M companies, enrich contacts, score leads, scrape websites. Growth: manage pipelines, send emails from connected accounts, read inbox and calendar, generate growth plans. Service: deploy widgets, monitor conversations, manage tenants. Automation: spawn sub-agents, schedule jobs, store facts, send team notifications. Every tool connects to live production systems.

02

Autonomous Workflow

A single instruction triggers multi-step execution. 'Find SaaS companies in DACH and start outreach to their CROs' becomes: search Index, score companies, research contacts, draft personalized emails, queue for approval, send via connected email, create pipeline deals, schedule follow-ups. Orchestrated by the agent state machine with full audit trail.

03

Multi-Agent Execution

Badger spawns child agents for parallel processing. Research ten companies simultaneously. Draft outreach for multiple prospects at once. Three levels deep, five concurrent workers. The parent continues working while children execute in the background. Every sub-agent has its own state machine and conversation history.

Architecture

01

Agent State Machine

Formal lifecycle tracking: idle, initializing, thinking, tool calls, validation, completion. Every state transition persisted. Subagents spawn up to three levels deep. Full audit trail of every AI decision.

02

Multi-Model Orchestration

Multiple frontier AI models routed by task type. Auto-classification selects the optimal model based on reasoning depth, speed, and cost. Dual streaming with token-by-token delivery across all model providers.

03

Dynamic Tool Extensibility

Open tool protocol enables dynamic tool discovery at runtime. Connect external tool servers via settings or conversation. Badger auto-discovers tools, namespaces them per tenant, and integrates them without code changes. Unlimited extensibility.

04

Memory Architecture

Short-term memory for active sessions with automatic expiry. Long-term memory for facts, conversation history, and embeddings. AI-driven context summarization compresses old messages. Coherent context across conversations spanning weeks.

05

48 Native Microservices

High-performance customer operations stack built for speed. Services handle real-time voice, chat, multi-model routing, and speech processing. Five service integrations: widget history, chat history, online clients, context management, personal data. Sub-second latency.

06

Email and Calendar

Your email and calendar, connected directly. Badger reads your inbox, searches emails, checks campaign replies, fetches calendar events, sends from your actual account. Preserves threads and signatures. Not from a generic address.

07

Kol Voice Engine

Proprietary speech pipeline built in-house by our co-founder. Recognition and neural voice synthesis developed from the ground up. Real-time sentiment analysis, speaker diarization. Natural phone and chat conversations between AI agents and customers.

08

Multi-Tenant Security

Row-level security at the database layer. Every piece of data scoped to a tenant. Isolated workspaces, agent configs, tool sets, knowledge bases. Tenant A never sees Tenant B's data. Zero shared state between organizations.

Research

Research-driven from day one.

Every product we ship is grounded in original research. Our team holds PhDs in artificial intelligence and has built foundational work in transfer learning, speech recognition, and agent architectures.

01

Agentic Systems

Multi-step autonomous reasoning, tool orchestration, error recovery, and long-running workflow management for enterprise-grade reliability.

02

AI Model Intelligence

Grounding AI models in structured business data. Multi-model orchestration, confidence calibration, and domain-specific optimization.

03

Speech and Voice AI

Real-time speech recognition, speaker diarization, neural text-to-speech, and sentiment analysis for natural business conversations.

04

Human-AI Collaboration

Trust-building agent behavior, feedback loops, oversight patterns, and measuring what makes work meaningful when routine tasks disappear.

Join Us

We're replacing how the world does commercial work.

Salesteq is a small team building something that hasn't existed before. AI agents that don't assist — they operate. We're looking for people who want to be part of that shift. Not job titles. Not perks lists. The chance to build the system that changes how companies run.

We don't have a careers page with open roles. We have a mission and we're looking for people who read this and feel something. Engineers, researchers, commercial minds — if you're convinced AI will fundamentally change how work gets done and you want to be on the team that builds it, reach out.

[email protected]Zug, Switzerland — or remote