Where the name comes from
Inspired by a 17th-century metaphor. Driven by 21st-century intelligence.
nuefunnel was born in Nuremberg, Germany — inspired by the Nuremberg Funnel (Nürnberger Trichter), a centuries-old metaphor for pouring knowledge directly into the mind. We loved that idea because it captures what we're trying to do with AI: make complex things feel simple.
One might call it context engineering or prompting skills, but ensuring that an AI solution has relevant knowledge about your business, processes and tools is the key to success. We build AI-native products and solutions that help businesses streamline processes, boost productivity, and unlock the full potential of modern AI.
Who we are
The two of us.
Sidharth Ramachandran
[Data science, ML]
Spent the last decade leading multiple data and product teams at RTL, Citibank and Transunion. Co-authored an O'Reilly book on Natural Language Processing. Started nuefunnel because most AI-product engagements are too large to be useful — a two-person team with coding agents can do in five weeks what a ten-person consultancy used to do in six months.
Adhisma Das
[Conversational AI, Partnerships & Consulting]
Led partnerships with Meta, Viber and Apple for conversational AI bots. Previously managed consulting assignments across EMEA and Latin America for a large consulting firm. Has a long-standing interest in the seam where conversational interfaces meet operational data. Joined to focus on what AI-native products actually look like when they're not glued onto an existing SaaS.
How we work
Small scope, two people, coding-agent-assisted.
We engage in short cycles.
Four to eight weeks per project. We don't do indefinite staff augmentation. The end of an engagement is a real end - you get a running system, the infrastructure-as-code, and the context to run it.
We write code with coding agents.
Both of us use coding agents heavily. It's what makes a two-person team viable at this scope. We're opinionated about which parts of the work agents help with and which parts they don't.
We don't introduce new platforms.
Almost every engagement runs on top of the customer's existing warehouse, CRM, or doc system. New systems-of-record are a ten-person-team kind of project - not something we take on.
We turn down work that isn't a fit.
Enterprise compliance work, ML-platform builds, and projects that need a project manager between us and the decision-makers aren't a good fit for the studio. We say so up front.
Get in touch
Send a short note.
We’re proudly bootstrapped, deeply curious, and aim to deliver value. Tell us what you have in mind and we'll reply within a couple of working days.