Frontend ownership and full-stack delivery
Sherlock AI
An AI-powered security product that analyzes smart contract repositories and pull requests, helping Web3 teams identify, review, and remediate vulnerabilities during development.
Product flow
From repository setup to actionable remediation
The product turned a technically dense security-analysis process into a workflow that engineering teams could understand and act on.
01
Repository setup
Connect a GitHub repository and choose the code or change to analyze.
02
Analysis run
Start and monitor an AI-assisted smart contract security analysis.
03
Findings review
Review potential vulnerabilities, severity, and supporting explanations.
04
Code context
Inspect the relevant code and attack path behind each finding.
05
Remediation
Use the finding context to understand and address the issue before release.
What the product does
Sherlock AI analyzes smart contract code during development so Web3 teams can identify risky logic and likely vulnerabilities earlier. It integrates with engineering workflows around repositories, commits, and pull requests, connecting findings to the code teams are preparing to release.
My role and operating context
I worked in a lean engineering team, collaborating directly with the CTO on engineering and the CEO on product direction. I owned all frontend architecture and implementation for Sherlock AI from the beginning, building the product with Next.js and TypeScript.
I collaborated closely with a designer and contributed product and interaction decisions as requirements evolved. Together, we turned a technically dense security workflow into a customer-facing product that teams could use throughout development.
The workflow I built
- Repository setup and integration with GitHub development workflows.
- Analysis-run states and progress across long-running security work.
- Findings lists and detailed vulnerability review.
- Code context and attack-path views tied to each finding.
- Remediation flows that helped teams understand and address issues.
Full-stack delivery
Although the frontend was my primary area of ownership, feature boundaries did not stop at the interface. I implemented supporting APIs and backend services, changed data models, and handled authentication work when the product required it.
That scope allowed me to deliver complete product workflows instead of treating frontend implementation as a handoff from other engineers.
Applied AI work
I integrated AI models into the analysis workflow and contributed to prompting and benchmarking. The product needed to turn model behavior and output into findings that users could inspect in context and use during remediation.
My work connected the AI-analysis layer to the product experience, including analysis states, structured findings, code context, and the interactions around reviewing results.
Public product evidence
Sherlock publicly presents Sherlock AI alongside protocols and ecosystems including Aave, Ethereum Foundation, Arbitrum, MegaETH, Sky.money, Optimism, Tempo, and Centrifuge. The testimonial below provides a direct example of product use.
“We’ve tried many different AI audit tools, and none come even close to Sherlock AI. Other tools are either good at PR runs or manual runs. Sherlock AI is by far the best combination of both.”