Isogeny — Tech With Integrity
Proof of Work

What We're Building

Isogeny designs, builds, and operates software products — starting with Hone, our own interview preparation platform. Here's a look at what we're working on, and the client work we'll add as projects complete.

Featured Project — Own Product
01

Hone

Your AI‑Powered Career Interview Companion

In DevelopmentWeb · iOS · Android

The Problem

Most candidates prepare for interviews in a fragmented, ad‑hoc way. Existing tools are either static question banks or narrow point solutions — none offer a genuinely personalised, end‑to‑end preparation experience.

  • No deep personalisation to your actual CV and target role
  • No tag-level insight into real strengths and weaknesses
  • Mock interviews that feel nothing like the real thing
  • Privacy and GDPR compliance treated as an afterthought

The Solution

Hone analyses your CV and job description to build a personalised preparation workspace — covering the specific skills, competencies, and topics your target role requires. A structured learning loop takes you from foundation knowledge to a full interview simulation, with tag-level analytics showing exactly where you stand at every stage.

  • Workspace personalised to your CV + JD in seconds
  • Dynamic Tag Cloud maps every skill you need to master
  • Two interview practice modes: turn-based and live voice
  • GDPR-first — personal data pseudonymised before any language model sees it

Who It's For

Graduates

Final-year students and new entrants preparing for their first professional roles

Professionals

Mid-career candidates seeking promotion, a lateral move, or a step up

Career Switchers

People moving into new industries or roles who need to build credibility fast

Key Features

CV + JD-Aware Onboarding

Users upload their CV and job description. Hone parses both alongside a free-form brain dump to generate a personalised preparation workspace in seconds.

Dynamic Tag Cloud

Every skill, competency, and concept is mapped to a tag. Tags update in real time as the user practises, driving adaptive recommendations and SWOT analytics.

Six Learning Modes

Study Resources, Flashcards, MCQs, Typed Answers, Code & Syntax, and Interview Simulation — a structured progression from foundation to full mock interview.

Adaptive Learning Loop

Scores below 70–80% route users back to weaker areas automatically. The system escalates question difficulty as mastery grows.

Walkie-Talkie Simulation

Turn-based interview practice using synthesised audio questions and voice or text responses. Session transcripts, tag scores, and structured feedback are retained.

Live Conversational Mode

Premium real-time two-way voice interview over WebSockets. Streaming speech recognition, language model responses, and speech synthesis combine for a natural, low-latency interaction.

SWOT Performance Dashboard

Tag-level heatmaps, strength/weakness breakdowns, deck completion metrics, and trend charts give candidates a clear picture of readiness over time.

GDPR-First Architecture

Personal data is pseudonymised before reaching any language model. Data minimisation, right to deletion, configurable retention, and audit logging are built in from day one.

Technical Architecture

Hone uses a layered architecture combining modern language models with conventional, well-tested web technologies.

LLM (general)Question generation, onboarding logic, and performance analysis
LLM (realtime)Cost-optimised streaming model for live interviews
RAGEmbeddings store for context-aware resource retrieval
STT / TTSCloud speech-to-text and text-to-speech for voice modes
BackendNode.js + TypeScript / Python + FastAPI
DatabasePostgreSQL + S3-compatible object storage
AuthOAuth2 / OpenID Connect (Google, GitHub)
RealtimeWebSockets + WebRTC for live interview mode

Pricing Model

Free

Core features with usage limits — a subset of preparation tiles and basic analytics.

Standard

Full onboarding, all preparation tiles, complete performance dashboard, and turn-based interview practice.

Premium

Higher usage limits, live voice interview mode (billed by minutes used), and priority processing.

Launch Roadmap

Alpha0–2 months

Core onboarding, tiles skeleton, basic turn-based interview mode

Closed Beta3–5 months

Full tiles, Tag Cloud, SWOT dashboard, initial pricing

Public Launch6–8 months

Free + Standard tiers live, early Premium beta

Phase Two9–12 months

Live voice interview mode, iOS + Android apps, role-specific content retrieval

Full case study, metrics, and demo to follow at public launch.

Interested in working together? →
Client Work

Workflow Automation & Integration

Enterprise client — case study in preparation

Mobile App — iOS Launch

UK startup — details to follow post-launch

Have a project in mind?

We'd love to hear about it — whether it's a new product, an internal tool, or an existing system that needs a fresh approach. No sales pitch, just an honest conversation.

Start the conversation