Please tell us a little about what TwoWay’s solutions are designed to do and the problems it is setting out to address.
TwoWay started with a first conversation with an FX Option trader. A market maker at a top-tier bank who shared his problem: receiving hundreds of broker chats every hour, missing up to 80% of that, and struggling to keep his vol well marked to market. He shared stories of times when the broker market would shift, with no market data platform reflecting it for hours. And during that time, if you don’t check your chats, you keep distributing wrong prices to your clients. Since then, we have had 100+ more interviews with traders across FX, rates, and fixed income, all sharing the same pain: an abundance of unstructured data in the interbank market that traders handle manually to update their curves.
One of our partners, a top-tier UK bank, recently shared an average estimate of that problem for them: £5K per trader per day, lost in mispricing and missed opportunities. That’s £1.2M per trader per year. TwoWay solves that challenge. Our solution ingests broker communications in real time, parses them with proprietary models, and surfaces structured market intelligence in milliseconds, giving voice traders real-time market awareness and giving e-traders more accurate curves to power their algos and distribution. And we do it across FX Options, FX Swaps, NDFs, and Fixed Income.
What types of firms are using TwoWay and is it mainly targeted at larger institutions?
Our first clients and partners are top-tier financial institutions. Some of them have already invested in building their own algos to parse the chats. They’re familiar with that pain point and understand the difference between building a prototype at 80% accuracy and getting an enterprise-grade platform to production, with the confidence that the prices feeding your curves are accurate. Others are new to that challenge.
They generally chose TwoWay for two reasons: (1) to increase their confidence in the prices extracted from the chats to feed their curves automatically, thanks to our algorithm and data quality layers; (2) to fit traders with a purpose-built interface that gives them genuine market awareness and context, not another dashboard to manage.
Beyond Tier 1 banks, we see significant interest across tier 2 and tier 3 institutions, where the problem is the same, the resources to solve it in-house are limited, and the pressure for efficiency is equally strong.
What are some of the key features and functionality your platform offers?
In simple terms, TwoWay first turns chats into structured data in real time, then makes sense of it.
Our proprietary models assign a confidence score to every parsed message, maintaining high accuracy while ensuring data quality and control when automatically feeding the traders’ curves.
Then TwoWay compares broker prices with the trader’s own marks to provide the right insights, highlighting pricing risks or market opportunities. In addition to real-time analytics for traders, TwoWay builds the data foundation for broker market data to feed risk and IPV systems.
How have you engineered TwoWay to take advantage of powerful new technologies like AI?
TwoWay is built using a hybrid approach that combines deterministic algorithms with AI models. The right balance between the former and the latter ensures we can get results in milliseconds while handling ambiguous context-dependent language or verifying and enriching the output. This approach allows us to achieve both high accuracy and millisecond performance, both of which are critical to traders.
While our objective is to take on as much of the traders’ tedious work as possible, we also ensure traders retain control and can review everything the app does at any time. That’s why we’ve built TwoWay with a traders’ feedback loop by design, allowing the algos to improve continuously from real desk corrections, and with a data control and quality layer.
What about performance? How do you ensure zero-latency data processing to enhance decision-making and pricing accuracy?
Speed is non-negotiable in trading, so our architecture is engineered for millisecond display with progressive enhancement: traders see parsed data as messages arrive in real-time, not after a batch process. The hybrid architecture described above enables this. Deterministic algorithms return results in milliseconds for more than 90% of messages, with AI enrichment layered on top.
Compliance and security are other critical requirements for institutional trading firms. How do you go about ensuring these essential needs are met?
We treat security and compliance as core to our product, not an afterthought. Our deployment architecture is built to meet institutional-grade standards. We offer our clients on-premises deployment, so they can keep control of their data, and none of it leaves the bank’s servers.
TwoWay’s data quality and control layer is designed to satisfy the data governance requirements of regulated financial entities, including the creation of a complete audit trail of what the market showed, and a full history of how traders engaged with that data.
What are some of the main benefits financial institutions are getting from using TwoWay?
TwoWay delivers value at three levels:
- Level 1 is direct P&L protection: eliminating the mispricing and missed opportunities that cost desks over $1.5M per trader per year.
- Level 2 is franchise impact. Once you are pricing with confidence and capturing the full broker flow, you can offer tighter prices to clients, improving your hit ratio, reputation, and franchise.
- Level 3 is efficiency. Our clients report 20–30% efficiency gains for their traders. In FX Options, that means responding to client requests 10x faster; in FX Swaps, managing up to 10 currency pairs instead of 3-4. Both compound significantly during market stress, when staying on top of your book is what separates the best desks from the rest.
How does your approach to product development differ from many other traditional fintech companies?
The difference is simple: we build with traders, not for them. Every feature has been developed alongside front-office FX and FI traders under real market conditions. And it’s still the case in production, iterating as the market moves. Our product decisions are grounded in what genuinely changes a trader’s day, not what looks impressive in a demo. As an AI-native company, we also ship fast. We iterate in days and weeks, not quarters. When a trader identifies a gap, we close it in the next release cycle.
How have you gone about making integration with TwoWay as seamless as possible?
We designed TwoWay to minimize integration friction wherever possible. The platform is delivered as a containerized, cloud-ready deployment that sits alongside existing systems. It doesn’t replace a bank’s parsers or pricers but enriches them. We connect to chat platforms’ APIs through pre-built connectors and provide hands-on support for any additional integration work. The average deployment and integration project takes 2 to 3 weeks.
TwoWay is a ready-to-deploy solution. What advantages does that bring?
A ready-to-deploy solution means the path from decision to live traders is weeks: 2–3 on average. That matters because every week faster is a week of value captured: at $1.5M per trader per year, time to deployment has a very measurable impact.Building this capability in-house is the alternative some institutions consider, but what looks like a 6-month project consistently takes 2+ years to reach production quality, if it gets there at all. A ready-to-deploy solution skips all that, accelerates time-to-value, and reduces project risk.
You offer a purpose-built user interface. Why is that important?
A trading desk is a high-pressure environment where every second counts and cognitive load matters. A generic analytics interface is not fit for that context, and traders are quick to abandon tools that don’t match how they actually work. TwoWay’s interface was shaped through constant iteration with traders. The information hierarchy reflects trading priorities, not software conventions: the most actionable signal, a mismark, an opportunity, or an unusual broker move is made to surface at the right moment, without the trader having to dig for it. The result is a genuine sense of the live market at a glance. The proof is adoption. As one of our traders puts it: “Not having TwoWay would be like driving in heavy rain without windscreen wipers.” That’s what turns a 2-week test into a permanent fixture on the desk.
What options do clients have for testing TwoWay before buying or deploying it?
We offer a structured evaluation pathway that aligns with how institutions actually make technology decisions, with flexibility in the level of pre-deployment testing required. TwoWay is production-proven across top-tier FX desks, enabling new clients to go from decision to live deployment faster than a traditional PoC cycle would allow.
For institutions that want a more gradual ramp-up, we’re happy to accommodate that too. The path is shaped around what gives each client the confidence they need to move forward. Either way, we work closely with each institution’s team throughout, and the goal is always the same: demonstrating real value as quickly as possible.
How can firms looking to use TwoWay or collaborate with you find out more information about doing that?
You can find more information on our website at www.twoway.finance, reach out to contact@twoway.finance, or connect directly with Guillaume Spay or me on LinkedIn. We always welcome a good conversation and are keen to collaborate and find ways to help.