Technical fluency
I do not write production code. I run the teams that do. I understand the work well enough to scope it, prioritize it, test it, and take it through audits.
3 yrsrunning engineering scrums
15technical proposals scoped with engineers
4external engineering labs directed on the what and why
12public projects shipped as product lead, 2023 to 2026
99.99%+uptime under my incident response
Problems solved on-chain at Charli3
PROBLEMCardano's 20-second block time was too slow for the data our customers needed.
WHAT WE DIDWorked with the team to build a layer 2 partnerchain on Substrate with 6-second blocks. Built a custom bridge that brings that data back to Cardano while keeping Cardano's security guarantees.
PROBLEMReferencing a single price feed on-chain was expensive.
WHAT WE DIDBundled price feeds into one reference using the EUTXO model.
PROBLEMProof of reserve on a push oracle lacked the precision and control a bank customer requires.
WHAT WE DIDMoved the service to a pull oracle so the customer can request a reserve update on demand. Along the way we solved privacy requirements on bank endpoints, rate limits, mismatched update timelines, and different datum structures for the output.
PROBLEMPrice APIs from DEXs and aggregators expose an oracle network to stale data and corrupt endpoints.
WHAT WE DIDThe team designed and shipped an MIT-licensed Python DEX aggregator. It pulls prices directly from liquidity pools, weighted precisely, and accounts for transfers between pools before each block update.
PROBLEMTraditional oracle networks do not validate off-chain data. Some use cases require it.
WHAT WE DIDBuilt an oracle validation service. Nodes check a trusted endpoint for a value, compare it to the on-chain value, and return agreement or a replacement once consensus is reached. Shipped an end-to-end proof of concept with Adobe Magento that validated e-commerce coupon codes.
PROBLEMProtocols wanted to call price feeds inside their own user flows, including payment.
WHAT WE DIDThe team created the pull oracle to meet that need.
PROBLEMHard forks broke things, and some problems were beyond the team.
WHAT WE DIDBrought in outside expertise to augment the team. Recruited several high-profile Cardano technical leaders to support projects where my own skills fell short.
How I ran the product function
- Took over the product team when the CTO left in 2023 and ran it through 2026. Shipped 12 public projects, each with a detailed scope.
- Created the engineering projects and workflows. Broke work into tasks and assigned them to team members.
- Wrote 15 technical proposals, each scoped with the engineering team.
- Ran quality testing before releases.
- Managed the node operator list and led support for node software issues.
- Working understanding of the major Web3 protocol types: DEXs, lending, stablecoins, prediction markets.
- Led the incident response team and wrote the incident reports. Borrowed the processes and standards from my year inside a tier-one bank.
- Primary lead on the 2024 on-chain, off-chain, and security audit. Took the findings back to the team and closed them, including complex security issues.
- Main non-technical lead on the CertiK audit.
- Built a financial audit team for a corporate rollover and led them through books that held only digital assets, one of the first audits of its kind in Canada.
- Team shipped the first EUTXO-based push oracle, the first pull oracle, one of the first active smart contracts on the chain, and the Python DEX aggregator.
Finn.ai, acquired by Glia
- Worked directly with engineering stakeholders on security and integration flows, model training requirements for new banking products, and launch requirements.
- Main contact for multiple solutions architects on the auth token framework that lets conversational AI exchange banking information over WhatsApp.
- Turned the bank's engineering needs into Jira tickets for the data science, model, and engineering teams. Tracked dependencies across teams to keep launches on schedule.
- Ran QA before each production launch. Troubleshot engineering issues with the team at 3am.
- Main contact on the account for technical issues. Understood the problem under pressure and got it to the right executive or engineering team.
Conik
- Translated the architecture of an AI data pipeline into investor slides a non-engineer could follow.
- Designed an end-to-end, self-improving infrastructure tied to AI oracles.
Predictable Revenue
- Comfortable in the room with CTOs, architects, and product managers.
- Can see a product at the technical level and at the level of detail a buyer cares about.
Tooling: AWS, Google Cloud, Vercel, Auth0, GitHub, Confluence, Jira, Notion, Figma. I use multiple frontier LLMs to build agents that improve my own workflow.
