My 4.5 years as a quant dev

Career Entry and Transition into Software

While studying Commerce at Queen’s University, Coding Jesus (CJ) won a national investment-banking competition but felt poorly suited to the industry’s culture. Rather than wait for conventional recruiting, he contacted BitMEX’s research team directly on Twitter and earned an interview by preparing a presentation on a topic he cared about.

At BitMEX, he began automating routine work with Q/KDB, built an internal database tracking the exchange’s identifiable blockchain addresses and transactions, and took paid Q/KDB courses. When management declined to move him into a programming role, he resigned.

He then studied C++ independently for 7–8 hours a day, building a small game and learning about programming and databases. Feedback from a friend helped him identify knowledge gaps and keep improving.

Interview Preparation

For LeetCode practice, he set a time limit: if an easy problem took more than 30 minutes or a medium problem more than 40 minutes, he would look up and study the solution. He then revisited the underlying concept and solved similar problems. He also used mock interviews to practice explaining his approach and choosing clear variable names.

He approached firms directly, including one that had no advertised opening. In interviews, he demonstrated perseverance through his self-directed study and Jane Street puzzle results—he says he solved 9 of 10 puzzles that year—and initiative by showing a C++ game he had built.

He recommends preparing projects so you can explain what you built, why you built it, what went wrong, and what you would change.

He received offers from SoundHound and a Chicago trading firm, choosing the latter because it better matched his goal of building a career in the United States. He argues that candidates should not automatically rule themselves out for lacking a computer science degree, a PhD, or visa sponsorship.

Quant-Development Work and Operational Impact

His responsibilities grew from junior software tasks into work on trading, market-data, and risk systems:

• Developed an RFQ-feed parser to interpret exchange data and classify structured instruments, then built tools that allowed traders to create custom CME spreads.
• Worked on latency measurement by matching market-data packets to order-entry requests, calculating timings from network-adapter timestamps, and sending results to Grafana through Elastic Stack and InfluxDB.
• Managed an application-monitoring and risk-control system connected to order-entry applications. For example, it could block automated trading gateways if a trade-reconciliation feed stopped signaling activity.
• Built a market-data pipeline for researchers. The previous process relied on slow Parquet files on network drives; the new pipeline reduced the time needed to obtain backtest data by 90%, enabling faster research iteration.
• Created market-data tools to classify order flow and interfaces for options pricing. To keep high-volume displays responsive, he subscribed only to data for options visible in the trader’s interface and unsubscribed as users scrolled.
• Later worked on direct-market-access and automated market-making systems, including hedging logic and trader-facing controls.

He emphasizes rigorous testing, low-latency design, and limiting interfaces to the minimum data required (“thin clients”), because faulty order-entry systems can cause major losses.

Management and Leadership

Technical skill was only part of the job. He also coordinated across time zones, communicated with traders and researchers, planned work, and managed stakeholder expectations.

He supervised two groups of interns working on exchange-simulator projects:

• One group added support for CME’s binary message format alongside the internal Protobuf/gRPC protocol.
• Another added support for CME MDP3 market data.

He also mentored two junior developers through daily meetings, task breakdowns, project assignments, and progress reporting. In addition, he interviewed candidates and made internship hiring decisions.

One difficult performance-management experience involved telling an underperforming intern directly that the company would not offer them a job. His lesson was that managers should communicate shortcomings clearly and respectfully rather than delegate every difficult conversation to HR.

From Quant Employment to Entrepreneurship

CJ stopped posting on YouTube for roughly a year and a half after his employer warned him that a video had crossed a confidentiality boundary. A 2023 interview process later reinforced that some firms viewed his public profile and résumé-screening business as a conflict. He broadly characterizes firms as either more social-media-friendly “new guard” organizations or more conservative “old guard” ones.

In 2024, he received offers from several quant firms but chose to pursue Get Cracked, motivated by the impact he felt he had made mentoring people and preparing them for interviews.

After leaving his employer in November 2024, he initially planned to focus full-time on the business after garden leave. In May 2025, anxiety about losing stable employment led him to look for a job. He accepted a trading-technology role at a Fortune 100 company with a reported salary of about $300,000, but resigned after two months once he regained confidence in the business.

He moved from Chicago to the UAE and focused full-time on expanding Get Cracked. His work included publishing content, adding learning paths and tests, building relationships with external recruiters, and hiring four full-time employees.

He also invested in hardware-engineering interview-preparation content, seeing a gap in the market and aiming to build a community around material he believed was not available elsewhere.

Get Cracked Elite

Get Cracked Elite is a 90-day mentorship program for aspiring quant developers, traders, and researchers. It includes:

• Industry practitioners as coaches
• Private one-on-one coaching and group meetings
• A master project and personalized learning roadmap
• Accountability check-ins
• LinkedIn positioning and résumé rewrites
• Referrals to external recruiting partners to support job placement

Key Metrics and Timelines

• 4.5 years: Tenure in quant development
• 7–8 hours per day: Self-directed C++ study after leaving BitMEX
• 30 minutes / 40 minutes: Time limits for struggling with easy / medium LeetCode questions before studying solutions
• 9 of 10: Jane Street puzzles reportedly solved in one year
• 90%: Reduction in researcher data-access time after building a new pipeline
• About $300,000: Reported salary for the Fortune 100 trading-technology offer
• Two months: Time before leaving that role
• Four employees: Full-time team hired for Get Cracked
• 90 days: Duration of Get Cracked Elite

Practical Takeaways

• Create opportunities by contacting employers directly and showing relevant work instead of relying only on posted vacancies.
• Use a structured study loop: time-box difficult problems, study the underlying concept, then practice it again.
• Make projects interview-ready by explaining their purpose, trade-offs, mistakes, and lessons.
• Treat data access, latency, risk controls, and testing as core business requirements in trading systems—not just implementation details.
• Invest in communication and people management alongside technical skills.
• Validate a business direction through a clear market gap, useful content, relationships with recruiting partners, and a structured service offering.

Presenters and Sources

• Presenter: Coding Jesus (CJ)
• Other source mentioned: Tech With Tim, who encouraged him to visit the UAE
• CJ also refers to unnamed friends and colleagues at BitMEX, Akuna Capital, Meta, and other trading firms.