Dishaya / Careers / Machine Learning Engineer

Machine Learning Engineer

Turn outcome data into a product that improves every week, without ever training on user content.

Engineering · Remote-First · Founding Equity

About Dishaya

Dishaya makes research people can actually send. One question in; a client-ready report, deck, and source ledger out, with every claim checked against the exact passage it cites and labeled verified, partial, unverified, or contradicted. What fails the check is disclosed, never hidden. We are early, independent, honest about both, and building toward one ten-year outcome: "Verified by Dishaya" becoming a mark a reader trusts before they read. More in About and Principles.

Why This Role Matters

Dishaya makes two promises that most companies treat as opposed: the product gets measurably better every week, and we never train on your content. This role exists to prove both can be true at once. You own the measurable quality of AI-assisted outcomes and the improvement loop behind it, built entirely on content-free outcome signals. If quality climbs month over month while the commitment holds without exception, that is your work, and it is the strongest trust argument the company can make.

What You'll Work On

• The measurable quality of AI-assisted outcomes: defining what better means, measuring it honestly, and moving it.
• The improvement loop: turning content-free outcome signals into changes that make next week's answers better than this week's.
• The no-training commitment, proven in practice: the product improves and user content is never the fuel.
• Cost and quality tradeoffs: the same rigor for less, or more rigor for the same, work package by work package.

We describe work by the outcomes you will own rather than by our internal systems; you will see everything on the inside from day one.

Responsibilities

• Own the definitions of quality for AI-assisted outcomes, and defend them with real measurements, not vibes.
• Build and operate the improvement loop end to end, from signal to shipped change to measured effect.
• Run cost and quality tradeoffs deliberately, and write down what you chose and why.
• Treat the no-training-on-user-content commitment as a design constraint you build within, never one you argue with.
• Write things down: decisions, reasons, and honest post-mortems.

Required Qualifications

• You have shipped machine learning in production and can walk through what broke, how you found it, and what you did about it.
• Rigorous evaluation habits: you design the measurement before the feature, and you distrust your own demos.
• Strong Python or TypeScript, and comfort owning a service in production: queues, retries, failure modes, observability.
• Clear, calm written communication.

Preferred Qualifications

• You have improved a product using aggregate or privacy-preserving signals rather than raw user data, and can explain what that discipline cost and bought.
• You have run cost and quality tradeoffs at scale and can reason about unit economics.
• Experience with retrieval, grounding, or citation-backed generation in a product users paid for.

Nice To Have

• Published writing, talks, or open-source work that shows how you think.
• Experience in a research-adjacent product: search, knowledge tools, document intelligence.

What Success Looks Like

• 30 days: you can state, in numbers we agree on, how good the product is today, and you have found the first gap between what we believe about quality and what we can actually prove.
• 90 days: one quality measurement you own has improved month over month, you can show the specific change that caused it, and the improvement came from content-free signals alone.
• 365 days: quality improving month over month is routine rather than an event, every improvement is provable, and the no-training commitment has held without a single exception on your watch.

Team Principles

• Honesty over fluency, in the product and in code review.
• Delete before you add; every abstraction earns its keep.
• Evidence over enthusiasm; direction comes from users.
• Small, senior, trusted; you own outcomes, not tickets.

Benefits

• Founding-level equity; early means it matters.
• Remote-first, judgment over time zones.
• The hardware and tools you need, without a procurement dance.
• Direct access to how the company runs: numbers, decisions, reasons.

Interview Process

1. Intro conversation (30 minutes): the honest state of the company, and what you want to build.
2. Craft deep-dive: real decisions inside work you shipped.
3. Paid working session: scoped, close to the real job, never spec work we ship.
4. References and a clear written offer, fast.

Equal Opportunity

Dishaya is an equal opportunity employer. We evaluate candidates on craft, judgment, and alignment with how we work, never on race, color, religion, gender, gender identity or expression, sexual orientation, national origin, disability, age, or veteran status.