High-Frequency Trading (HFT) sits at the intersection of markets, maths, and extreme technology. In this Q&A, Rajib Ranjan Borah, Co-founder & CEO of iRage and Co-founder & Director at QuantInsti, breaks down what it’s really like to work in HFT, the skills you need, and how to build a career in one of the most competitive areas of finance.
About Rajib Ranjan Borah
Rajib Ranjan Borah is the Co-founder and CEO of iRage, a firm specializing in high-frequency and algorithmic trading across asset classes like equities, currencies, and commodities. Under his leadership, iRage has become one of India’s leading high-frequency trading firms. Rajib is also the Co-founder and Director of QuantInsti, where he has been instrumental in creating educational programs like EPAT, which trains professionals in algorithmic trading worldwide.
Read more here.
This Q&A is adapted from his session on breaking into HFT and how modern quant teams actually work on the inside.
Why did you get into HFT?
Rajib Ranjan Borah: "When you're doing high-frequency trading, you're competing with some of the smartest people on earth. And then, it's like playing a video game every day. At the end of the day, you get a score—you know if you’re doing well or not. That’s what makes it fun."
What is the work culture like in HFT?
Rajib Ranjan Borah: "HFT is like a Western movie where you have Clint Eastwood and then a lot of other actors. The industry is dominated by a few key players, with the winner taking a significant portion of the pie. It's a people business. The kind of talent that you hire defines how you will do. We focus on hiring people who have shown excellence prior to joining us."
How does HFT work?
Rajib Ranjan Borah: "You either make a prediction for a very short duration or find an opportunity that might not persist for long. You get hundreds of gigabytes of data per day, and during market activity, the microbursts are significant. The challenge is processing this information quickly and effectively to capitalize on market opportunities."
Watch next:
What skills are essential to succeed in HFT?
Rajib Ranjan Borah: "The traditional boundary between developer, researcher, and trader is becoming blurred. The successful ones do the research, implement it, trade it, get live feedback, and make changes. It’s a mix of quantitative analysis, coding, and financial markets expertise."
In modern HFT teams, the strongest people own the full loop: they generate ideas, test them, implement them in code, monitor live performance, and iterate.
Skill stack in one view:
- Quantitative thinking – probability, statistics, optimisation, comfort with noisy data.
- Programming – writing clean, efficient code; working with data structures and performance.
- Market understanding – order types, microstructure, liquidity, and how different participants behave.
- End-to-end ownership – not just theory or code in isolation, but the whole research → deploy → debug cycle.
Deep-dive on roles:
Should I learn Python or C++ for HFT?
Rajib Ranjan Borah: "To quickly make insights and test hypotheses, people use Python. But as the amount of data they analyze grows, they start using C++. The execution part, especially if it's low latency, must be in C++. If you're a developer working on exchange connectivity and low-latency risk management, C++ is essential."
How to approach this as a learner:
- Start with Python to build intuition, analyse data and prototype strategies quickly.
- Add C++ as you move into low-latency or infrastructure-heavy roles, or when your research starts hitting performance bottlenecks.
Most serious HFT professionals eventually become comfortable in both.
Compare languages:
How important is programming in HFT?
Rajib Ranjan Borah: "Programming is not as scary as it is made out to be. It's just logical thinking. But if you're aiming for a meaningful role in a top HFT firm, you have to be good at programming. You have to code your strategies, test new hypotheses, and adapt in real-time."
You don’t need to be a “rockstar coder,” but you do need to be comfortable writing and reading code daily. That means:
- Automating your analysis instead of doing everything by hand.
- Implementing and debugging your own ideas.
- Handling large datasets and performance issues without panicking.
If you avoid coding altogether, you’ll be limited to very narrow roles in this industry.
What courses or certifications help in getting into HFT?
Rajib Ranjan Borah: "CFA is a fundamental course but not stats-heavy or quant-heavy. CQF is quant-heavy but not necessarily focused on capital markets. EPAT (Executive Programme in Algorithmic Trading) was designed by people from an HFT background and is well suited for systematic trading and HFT roles."
What qualities help in cracking HFT interviews?
Rajib Ranjan Borah: "Perseverance is key. You might give up just before becoming Clint Eastwood. We look for ‘sparks’ in resumes. We check for your passions, hobbies, and interests. Have you pursued something to a high level of excellence? HFT requires deep thinking, so we look for people who dig deep into problems rather than just scratching the surface."
How to signal “spark” on your resume:
- Highlight 2–3 projects where you went unusually deep (competitions, research, tools, open-source work).
- Show measurable outcomes: competition ranks, P&L on paper-traded strategies, performance benchmarks of your code.
- Be ready to talk about failed ideas and how you iterated, not just your wins.
The common thread: curiosity plus stamina.
You can check EPAT Projects by our Alumni here:
Featured Work You Can Replicate
What is an EPAT Project?
The Executive Programme in Algorithmic Trading (EPAT) is built for professionals who want to turn trading ideas into strategies. At the heart of the programme is the Algo Trading Project a hands-on project where each participant builds, tests, and documents a trading strategy using real market data.
Is machine learning used in HFT?
Rajib Ranjan Borah: "Yes, we use ML techniques in multiple cases for research. You have to reduce the universe of features to a workable set. While ML is not always used directly for trading strategies, it plays a key role in data analysis and pattern recognition."
In many HFT firms, ML is used heavily in research, even when the final live trading logic is intentionally kept simpler for robustness.
Typical uses:
- Reducing thousands of potential features down to a manageable set.
- Detecting regimes where certain strategies work better or worse.
- Spotting anomalies in fills, latency, or slippage that may indicate infrastructure or market issues.
Watch the following video as Rajib elaborates on how AI, machine learning, and new-age technologies are shaping quant jobs and high-frequency trading. Understand what skills and tools are essential for staying competitive in algo trading.
What advice do you have for aspiring HFT professionals?
Rajib Ranjan Borah: "Do things that you enjoy. Get into an industry that resonates with what excites you. If competing with the smartest people and seeing a daily score excites you, get into HFT. If not, find what truly excites you."
How do SEBI regulations impact HFT?
Rajib Ranjan Borah: "There are two sets of rules—one for retail and one for index trading. These changes mean the market microstructure will change, which is both a threat and an opportunity. If you are fast and nimble, you will find opportunities because trading styles will change. Traders who can identify how behavior shifts will benefit from regulatory volatility."
How can someone secure a quant role in HFT?
Rajib Ranjan Borah: "Try to participate in programming contests, publish your code on GitHub, and join hackathons. Use platforms like Quantra or BlueShift to learn about backtesting and quantitative trading strategies. Hands-on experience is key."
Editor's Notes: Immediate Practical Steps
Focus on these concrete actions to build practical skills:
- Engage in Competitions: Participate in coding or quant finance challenges.
- Build and Share: Publish small, documented projects/strategies on GitHub.
- Learn and Practice: Use Quantra for concepts and Blueshift for real-market backtesting.
- Critique Your Work: Analyze strategy performance (Sharpe, drawdown, turnover) to improve models.
Consistent execution of these steps creates a development portfolio demonstrating end-to-end strategy execution.
What is the difference between scalping and market making?
Rajib Ranjan Borah: "Market making is providing liquidity, while market taking involves short-duration price predictions. Arbitrage is finding mispricings and locking into those prices. Scalping involves capturing small price inefficiencies for profit."
What kind of hardware is used in HFT?
Rajib Ranjan Borah: "GPUs enable you to do a lot of things in parallel. FPGAs process data at the hardware level, reducing latency. Almost everyone in HFT now uses FPGA programming to minimize delays. The game has moved from software optimization to hardware optimization."
Final Thoughts
Rajib Ranjan Borah: "Be true to yourself and aim for excellence in whatever you do, whether in HFT or another industry. Never get into an industry where coming to work feels like a chore. You have to spend 30+ years in your career—choose something you love."
Next steps
Watch the full webinar here:
Build skills with Quantra
Develop practical skills in Python, ML, and backtesting with these courses:
- Python for Trading: Basic – Beginner Python for markets (data structures, core libraries).
- Introduction to Machine Learning and AI for Trading – Apply core ML concepts and algorithms to market data.
- Backtesting Trading Strategies – Design, test, and evaluate trading strategies realistically.
- Day Trading Strategies for Beginners – Momentum, scalping, and HFT intraday ideas, coded and tested.
Structured learning with EPAT
For an intensive path:
- Explore the EPAT – Executive Programme in Algorithmic Trading for full details.
- Schedule a call with an EPAT career counsellor for career alignment advice.
Experiment on Blueshift
- Use Blueshift for research, backtesting, and paper-trading with institutional-grade data, building an interview-ready portfolio.