The new AI Central

We've taken a lot of learnings from our launch in 2024 to bring a set of features which add tremendous value to Human centered AI workflows in companies across sizes.

Let's take the example of a 1-1 ABM workflow in a typical B2B marketing scenario. The following sections explain how running this in AI Central will look like

Custom AI enabled workflows

We're bringing AI agents, Custom Prompts and Human Review all together in a manner which mirrors the actual scenario in workplaces.

So , in this example of ABM, our pre-VALIDATED AI ABM agent will transform the 1-1 deep research-driven Account-Based Marketing (ABM) workflows by integrating models with context, tools, and human oversight for precise, personalized account engagement. AI agents orchestrate these steps, autonomously researching targets while handing off to humans for nuanced decisions.

The AI agent used pre-validated which means that it has been used for many clients but yes, we will need to set it up for your offering, tonality and workflow.​

External Tool Use

Tool use elevates AI from advisory chat to operational execution in ABM, invoking external tools like LinkedIn scrapers, CRM databases (e.g., Salesforce), intent data platforms (e.g., ZoomInfo), and email automation systems. AI agents invoke these as first-class actions: a research agent fetches technographics, a personalization agent drafts emails, and an outreach agent schedules messages—completing workflows end-to-end. Humans intervene only for strategic review, ensuring precision in high-stakes B2B pursuits. Things have changed. Today, you need not buy expensive SAAS software, you can simply use API credits for the exact feature that you are interested in. And that's what an AI enabled workflow will help you to do.

​

Dynamic Context, so important for Marketing and HR

Dynamic context pulls real-time account data via protocols, accessing firmographics, recent funding news, content consumption history, and competitive intel at runtime for hyper-personalized ABM plays. AI agents compose this context from APIs like Clearbit for stakeholder profiles or G2 for review analysis, enabling tailored proposals addressing specific pain points like "your recent CISSO hire signals cybersecurity gaps." This portability shifts moats to owning mappings between account signals, marketing policies, and decision triggers.

​

Intelligent AI Model Routers, bring on the specialists.

Intelligent routers select optimal models or agent sequences for ABM tasks based on needs—small models for quick intent scoring, reasoning models for synthesizing deep research reports, or specialist AI agents for multilingual content adaptation. For complex accounts, it chains agents: route to a research agent for org chart mapping, then a creative agent for persona-specific messaging, optimizing for latency in time-sensitive outreach. Over time, routers learn from patterns, like shortcutting simple email variants while escalating nuanced competitive positioning analysis. Using this results in good cost savings also in the long term and also you're using models for what they are best suited for.

​

System-Guided Prompting, not DIY.

System-guided prompting structures human input into optimized AI interactions, prompting people with targeted questions like "What key challenges does this account face?" or "Describe the ideal buyer persona," then auto-building comprehensive prompts incorporating those answers, dynamic context, and best-practice templates. This replaces DIY prompting in human-centric workflows, where agents refine responses—e.g., turning human notes into "Generate a pitch citing [user input] alongside [scraped competitor data]." AI agents enhance this by simulating multi-turn dialogues, ensuring prompts align with ABM goals like engagement lift.​

Telemetry

Telemetry tracks ABM system performance in production, logging route choices, task durations, engagement outcomes (e.g., open rates, meeting books), and human feedback loops. This observability enables automatic tuning, such as refining routers when low-intent accounts yield poor conversions or alerting humans to stalled pipelines. Metrics tie directly to ROI, like 208% revenue uplift from targeted campaigns.​