Your execution logs tell you what happened. We tell you why.
One-time diagnostic powered by a proprietary engine built over years of R&D on production trading systems. Every result is deterministic and auditable. Not AI-generated. Not a template.
Join the WaitlistWhat You Get
Built by engineers, not generated by AI. Every analysis is the product of years of R&D on live trading infrastructure.
Proprietary Engine
25+ analyses running simultaneously through a pipeline built over years of R&D on production trading systems. Every output is deterministic and auditable. Not AI-generated. Purpose-built tools, calibrated by engineers who have operated these systems.
Diagnostic Report
10-15 page PDF across 4 sections: Summary, Microstructure Context, Execution Analysis, and Recommendations. Accompanied by a 7-sheet XLS workbook with the raw data behind every finding.
Execution Health Dashboard
Fill rate, slippage, latency, reject rate, toxicity, and book imbalance at a glance. Color-coded. The first page of the report tells you where to look.
Top-3 Changes Memo
Two pages. Three findings ranked by dollar impact. Root cause, evidence, and a 30-day action plan for each. Deterministic results you can put in front of a board.
Per-Order Context
Every execution linked to the market state at that moment: toxicity, book pressure, spread, and depth. Not aggregates. The actual conditions around each order.
Actionable Guardrails
Specific thresholds you can implement: when to pull quotes, when to reduce size, when to widen spreads. Derived from your data, not from a textbook.
How It Works
Three steps. Three to five days. One complete picture.
Discovery
We learn about your operation, agree on scope, confirm your data format.
Analysis
Your data runs through the engine. Every finding is interpreted against real-world trading infrastructure context by our lead consultant.
Walkthrough
Full report delivered and reviewed live. You leave with quantified recommendations.
What We Analyze
The questions your execution data can answer – but isn’t.
Where is my slippage actually coming from?
You see aggregate slippage in your reports. We decompose it into timing cost, market impact, and sizing – so you know which lever to pull.
What was the market doing when I traded?
Toxic flow, book imbalance, liquidity depth, and resilience – captured at the exact moment each order executed. The context your execution logs can’t provide alone.
Are my orders behaving the way I think they are?
Reject spikes during volatile regimes. Cancel/replace chains that create exposure instead of reducing it. Queue positions that guarantee you’ll never get filled.
Is my speed actually earning me money?
Not all latency improvements translate into better fills. We measure where speed matters and where you’re over-investing in microseconds that don’t convert.
What’s the dollar breakdown of my execution losses?
Timeouts, partial fills, opportunity cost, fee drag – broken down so you can see exactly where dollars are leaking and what each fix is worth.
How is my venue actually performing?
Fill rates, reject patterns, reliability, slippage distribution – a complete scorecard for every venue you touch, not just the metrics your OMS shows you.
What a Diagnostic Reveals
Real patterns the engine detects. Each example maps to an actual report section with quantified findings.
A market maker was experiencing elevated adverse selection on passive fills, but only during specific 90-second windows following order book imbalance spikes. VPIN scores exceeded 0.85 during these windows while baseline sat at 0.4. The diagnostic identified that their quoting logic wasn’t pulling fast enough when toxicity spiked. Tightening the VPIN-based pull threshold eliminated the exposure.
1.4 bps improvement on passive fills
On $20M daily notional, that’s ~$70K/year. On $200M, ~$700K/year.
An algo desk had a 4% overall reject rate, which looked acceptable. The diagnostic disaggregated by market regime and found that during high event-density periods (>5K messages/sec), the reject rate spiked to 18%. These were exactly the moments when the strategy had its highest expected edge. The root cause was a rate limiter configured too conservatively for burst conditions.
2.1 bps of missed edge recovered
On $50M daily notional, that’s ~$260K/year. On $500M, ~$2.6M/year.
A stat-arb desk was investing in co-location upgrades to reduce round-trip latency from 45 to 28 microseconds. The latency-to-fill correlation analysis showed that below 35 microseconds, additional speed had zero incremental fill improvement on 3 of their 5 instruments. They were over-investing in speed on liquid names where queue position mattered more than raw latency.
Infra budget redirected to instruments where speed converts
Savings depend on current co-location spend. Typical reallocation: 20-40% of annual infra budget.
An execution desk was experiencing 3.2 bps average slippage on parent orders above $50K notional. The Implementation Shortfall decomposition revealed that 55% of the cost was timing slippage (delay between signal and first child order), not market impact. The signal-to-order pipeline had a 12-microsecond overhead from an unnecessary risk check that ran synchronously. Removing the bottleneck cut timing cost by 1.8 bps.
1.8 bps reduction in timing slippage
On $100M daily notional, that’s ~$450K/year. On $1B, ~$4.5M/year.
All examples from internal analysis using synthetic data. Impact scales with your trading volume. Each diagnostic engagement produces dozens of findings across 25+ analyses.
See what your execution logs are hiding.
We’re onboarding a limited number of firms. Join the waitlist and we’ll reach out to discuss your data.
No commitment. No sales pitch. Just a technical conversation about your data.
FAQ
At minimum: execution or order logs in FIX, JSON, or CSV. For the full diagnostic, we also need historical market data (L1/L2/L3) and public trades for the same period. We’ll confirm on the discovery call.
Signed NDA + Data Handling Agreement. Isolated encrypted workstation, accessed only by our lead consultant. Permanently deleted within 5 business days of delivery. No cloud, no third-party access.
3-5 business days from kickoff to report delivery.
Our lead consultant has 20+ years building production HFT systems. The proprietary engine was built from that experience. You work directly with a senior practitioner on both calls, not a junior analyst or a sales team.
You get the full execution-side analysis: order lifecycle, fill quality, slippage decomposition, venue performance, and the Top-3 Changes Memo with quantified dollar impact. That alone delivers actionable findings. Adding historical market data unlocks the microstructure context layer: VPIN toxicity, LOB imbalance, market resilience, spread dynamics, and cross-venue comparison. That’s the “missing 80%” that explains why your executions performed the way they did.
TCA is a report card. This is a playbook. 25+ simultaneous analyses through a proprietary engine, interpreted by someone who has built these systems. Standard TCA gives you neither the engine depth nor the expert context.
No. Every metric, decomposition, and finding is computed by proprietary analytical tools built over years of R&D on production trading systems. These are purpose-built engines, not LLM prompts or AI-generated summaries. We use AI openly in our daily workflow like any modern team, but the core analysis – the numbers, the decompositions, the microstructure measurements – comes from deterministic, auditable code written by engineers who have built and operated these systems. That’s the difference between a report you can act on and one you have to second-guess.
We’re currently in an early access phase, onboarding a limited number of firms to validate the service and refine the experience. Pricing will be finalized based on what we learn from these initial engagements. If you’re interested, join the waitlist and we’ll discuss scope and terms directly.