Build agents people can trust with real investment work.

default

To get started, replace the image opposite with a striking company photo to catch people's attention. Then, describe what makes your company great. Usually, an about us section will highlight some of the following: company history, company culture and values, future plans and goals, mission statement, description of products/services offered, awards and recognition received, customer testimonials or case studies and background.

About Us

Add title

Our Values

Add name

Meet Our Team

Gallery

<html><head></head><body><h3>About Itoflow</h3><p>Itoflow builds AI agents for investment teams. They research ideas, construct portfolios, assess risk, and help teams respond as markets change.</p><p>To earn trust, an agent must work from real data, use quantitative tools, and leave a clear record of what it did and why.</p><p>The founding team comes from investment banking, quantitative trading, and technology, including JPMorgan, RBC, Tower Research, Google, and Microsoft.</p><h3>Why now</h3><p>A prompt, tool, or model change can fix one task and break another. Without representative evaluations, we cannot tell.</p><p>We run evaluations throughout development, not at the end. We need realistic tasks, scoring, human review, and regression checks before we keep a change.</p><p>You will define how Itoflow measures and improves agents used for high-stakes investment decisions.</p><h3>The role</h3><p>We are looking for an AI Researcher to own Itoflow's evaluation and learning system.</p><p>You will build realistic tasks, scoring, and failure analysis across models, prompts, tools, data, and workflows. Where the evidence gives us a clear target, you will test reinforcement learning (RL) and reward-model approaches.</p><p>You will work with founders, engineers, product, and investment professionals. Your results will decide which agent changes reach users.</p><h3>What you will own</h3><ul><li>Build the evaluation suite for reasoning, tool use, source fidelity, decision quality, and regressions.</li><li>Create realistic tasks, baselines, scoring, and human-review loops that make agent changes comparable.</li><li>Test reward-model and RL approaches when they can improve a measured behaviour, and keep the results reproducible.</li><li>Trace failures to the model, prompt, tool, data, or workflow, then turn the findings into product priorities.</li><li>Decide when the evidence is strong enough to keep, change, or reject an agent behaviour.</li></ul><h3>What we are looking for</h3><ul><li>Experience bringing empirical AI research, agent evaluation, or applied machine learning into a real product.</li><li>Strong Python skills and sound experimental design, statistics, and data analysis.</li><li>Familiarity with RL, reward models, preference learning, or model evaluation methods.</li><li>Good judgement about measurement, baselines, human review, and the limits of automated scoring.</li><li>The ability to turn a broad question about agent quality into a testable plan and dependable implementation.</li></ul><h3>Compensation</h3><ul><li>Competitive salary based on experience</li><li>Early-stage stock options</li></ul></body></html>

<h3>About Itoflow</h3>
<p>Itoflow builds AI agents for investment teams. They research ideas, construct portfolios, assess risk, and help teams respond as markets change.</p>
<p>To earn trust, an agent must work from real data, use quantitative tools, and leave a clear record of what it did and why.</p>
<p>The founding team comes from investment banking, quantitative trading, and technology, including JPMorgan, RBC, Tower Research, Google, and Microsoft.</p>
<h3>Why now</h3>
<p>A prompt, tool, or model change can fix one task and break another. Without representative evaluations, we cannot tell.</p>
<p>We run evaluations throughout development, not at the end. We need realistic tasks, scoring, human review, and regression checks before we keep a change.</p>
<p>You will define how Itoflow measures and improves agents used for high-stakes investment decisions.</p>
<h3>The role</h3>
<p>We are looking for an AI Researcher to own Itoflow's evaluation and learning system.</p>
<p>You will build realistic tasks, scoring, and failure analysis across models, prompts, tools, data, and workflows. Where the evidence gives us a clear target, you will test reinforcement learning (RL) and reward-model approaches.</p>
<p>You will work with founders, engineers, product, and investment professionals. Your results will decide which agent changes reach users.</p>
<h3>What you will own</h3>
<ul>
<li>Build the evaluation suite for reasoning, tool use, source fidelity, decision quality, and regressions.</li>
<li>Create realistic tasks, baselines, scoring, and human-review loops that make agent changes comparable.</li>
<li>Test reward-model and RL approaches when they can improve a measured behaviour, and keep the results reproducible.</li>
<li>Trace failures to the model, prompt, tool, data, or workflow, then turn the findings into product priorities.</li>
<li>Decide when the evidence is strong enough to keep, change, or reject an agent behaviour.</li>
</ul>
<h3>What we are looking for</h3>
<ul>
<li>Experience bringing empirical AI research, agent evaluation, or applied machine learning into a real product.</li>
<li>Strong Python skills and sound experimental design, statistics, and data analysis.</li>
<li>Familiarity with RL, reward models, preference learning, or model evaluation methods.</li>
<li>Good judgement about measurement, baselines, human review, and the limits of automated scoring.</li>
<li>The ability to turn a broad question about agent quality into a testable plan and dependable implementation.</li>
</ul>
<h3>Compensation</h3>
<ul>
<li>Competitive salary based on experience</li>
<li>Early-stage stock options</li>
</ul>