Skip to Main Content

deepsense.ai

Global Codex Hackathon: [London]
Often there are projects that were created in the past, could still be useful, and would be more performant by running on a more modern stack. This still require human-in-the to keep the control over the process./…

Our project aims to create a set of hierarchical skills for migrating legacy Python code bases using Codex and subagents. By having more modular skills we are able to delegate work from a higher-level abstraction skill to more implementation oriented skills. The goal is to keep the same functionality and feel of the application whilst improving the underlying tech stack.

Team

pawel_kmieciktomekpscronberg

Placement

Participant

More from Global Codex Hackathon: [London]

Capco

Sentra is a real-time decision intelligence layer embedded in retail banking frontline interactions. It transforms fragmented customer data into a dynamic…

According to the FCA, over 50% of UK adults have at least 1 vulnerability characteristic. 13 million people still rely on physical branches / in person banking support. At the same time, retail banking agents currently operate with fragmented data, limited context, and increasing regulatory pressure, leading to inconsistent decisions, missed vulnerability signals, and potential customer harm. This solution solves that problem by providing real-time, explainable decision support, enabling agents to make consistent, compliant, and proactive customer-centric decisions./…

catalyst

In Common is an AI-powered workforce intelligence prototype that helps large organizations staff teams faster, more fairly, and with better evidence. It sits…

In large organizations, staffing often relies on incomplete profiles, keyword searches, and personal networks. That creates slow staffing cycles, missed opportunities, and unfair visibility for talent outside existing circles. In Common solves this by turning static workforce data into explainable, AI-driven talent intelligence that surfaces stronger matches, identifies skill gaps, and helps build teams based on capability, availability, and collaboration signals rather than proximity or guesswork./…

Cognita Reply

Our project combines an agent-driven development framework (AI-Led SDLC) with a plugin called LegacyBridge to enable interaction with legacy systems that lack…

Legacy systems are a major barrier to innovation across the tech industry. Many critical platforms lack APIs or modern integrations, making it difficult to access data, automate workflows, or build AI-powered solutions. Our project solves this by enabling agents to both understand and act within legacy systems, removing the dependency on APIs and overcoming the limitations of traditional RPA./…

Endavans

Our project is an AI-powered travel insurance claim processing platform that automates the full claims journey from document intake to final decision. It…

What we are solving, is the slow, manual, and inconsistent process of handling insurance claims. Today, claims teams often spend days or weeks reviewing documents, checking policy terms, validating evidence, screening for fraud, and coordinating next steps. We tackled this problem by building an end-to-end workflow that cuts manual effort, accelerates decisions, and brings full visibility into the claims process./…

Works On My Machine

Coachdex is a team intelligence layer on top of Codex. It turns isolated Codex sessions into shared team intelligence and reusable skills by capturing…

AI coding tools are creating value, but most of that value is trapped inside individual, ephemeral sessions. Developers each discover their own prompts, workflows, and habits, but teams cannot easily see what is working, standardise it, or reuse it. Coachdex solves this by converting ad hoc AI-assisted work into repeatable execution, measurable insights, and reusable institutional knowledge./…

CGI - HMCTS

An AI-powered software engineering assurance tool that turns a public GitHub repository into a structured review in minutes. A user submits a repo URL and the…

Software assurance reviews are slow, expensive, and inconsistent because they depend on manual evidence gathering and senior specialist time. In our client’s current model, a single review can take around five days end-to-end, and involves multiple senior reviewers and can result in bias. Our project compresses that process into an automated workflow that applies the same assessment structure every time, surfaces evidence directly from the repository, and gives decision-makers a faster starting point for governance and delivery discussions./…