- PhD Engineering
- MPhil Physics
PhD, University of Cambridge
Vision-capable agents that listen, see, think and act — so you can focus on the work only you can do.
Launching in private beta. No credit card. No spam — one email when access opens.
Voice, images and surrounding context are first-class inputs. Our agents perceive the moment — what was said, what is in the frame — before they reason about it.
Reasoning agents that plan, search, verify and decide before producing — never one-shot text generation. A detected object leads to the next automated step, not a wall of text.
You stay in control. Review, edit and approve. The agent does the work; you keep the voice — and the judgment — when it matters.

A voice-first tool for working professionals who want a steadier publishing rhythm.
Talk for two minutes. Live record or upload audio notes — whatever you already have.
Reasoning steps: extract your point of view, search for a fresh angle, draft a hook, structure the body, and include a call to action.
Review it side-by-side with the original. Approve, edit, or regenerate — then publish wherever your audience reads you.
“Most teams I talk to aren’t short on ideas — they’re short on throughput. Here’s what changed when we stopped drafting in text and started talking through the point of view first.”
Two founders with research backgrounds in Agentic AI-based Retrieval-Augmented Generation (RAG) Systems, Product Development Consultancy, and Digital Marketing — with a combined over two decades of experience building and marketing products that professionals actually use.
Short, occasional posts on how we design voice-first, vision-capable agents — and what we wish we had known sooner.

What changes when the keyboard becomes optional? A field note on latency, intent capture, recovery, and trust in voice-first systems.

Why we chose to start the conversation about Distl before the interface is finished — and what we hope to learn by treating polish as something a product earns, not something it ships with.
Coming soon
Insaltix Ltd is a UK product studio that designs multimodal Agentic AI products — software that listens, sees, thinks and acts. We build agents that perceive voice and vision, reason about context, and act on the next step while keeping a human in the loop.
Distl is Insaltix's first product. It is a voice-first tool that turns a short spoken reflection into a polished thought-leadership post. You talk for around two minutes, the agent drafts a structured post with hooks and image suggestions, and you review and publish.
Distl is opening in private beta in small waves. Register your interest on the home page and we will email you once a slot is ready. There is no marketing list and no follow-ups — one email when access opens.
No. Insaltix does not set any cookies and does not use non-essential trackers. If we introduce cookies in the future, we will update the cookie policy and ask for opt-in consent.
Insaltix Ltd is registered in England and Wales (Company No. 17123911) with its registered office at 20 Wenlock Road, London, N1 7GU, United Kingdom.
You can reach us by email at hello@insaltix.com or by post at the registered address above.
We’re opening private beta in small waves. Drop your email and we’ll send a single invite when a slot is ready — no marketing list, no follow-ups.