Source Job

• Build product surfaces across web, desktop, and mobile for Hermes Agent and Nous Portal.
• Develop backend services and APIs that hold up under real usage.
• Maintain the cloud platform and infrastructure behind products and model serving.

Nous Research builds AI infrastructure including Hermes Agent and Nous Portal. The team is small, fast-moving, and values generalist engineers with strong ownership and low ego.

• Build and improve core components of the Agent Runtime, including intent routing, RAG, and multi-step agent workflows.
• Design and optimize Tool Server / Tool Calling capabilities for agent execution and integration.
• Develop evaluation datasets and harnesses to measure agent quality and reliability.

Binance is a leading global blockchain ecosystem behind the world's largest cryptocurrency exchange by trading volume and registered users. With over 320 million users in 100+ countries, Binance offers a dynamic, flat-structured environment focused on innovation and growth.

• Own the shared AI foundation used by product engineering teams, including model selection, routing, context management, and evaluation.
• Design and ship agentic capabilities end to end, from proof-of-concept through production optimization and iterative improvement.
• Establish evaluation infrastructure and practices to measure model and agent performance, ensuring quality and accuracy for financial decision-making.

Our partner is building a suite of financial planning and analysis capabilities powered by a shared AI foundation. They operate as a remote-first engineering team with a focus on autonomy, craftsmanship, and high-impact work.

• Build and improve LLM-powered systems to navigate complex construction tendering workflows.
• Own features end-to-end from model/agent through API to TypeScript front-end.
• Collaborate across teams to translate business problems into shipped AI features and maintain production systems.

Cosuno is a fast-growing tech startup that revolutionizes construction through a digital platform for tenders and procurement. With a small, well-funded team, we foster a calm, collaborative, and productive engineering culture.

• Design, build, and ship AI-powered features end-to-end, from prototype to production.
• Work hands-on with LLMs, embeddings, RAG, and related tooling to solve real product problems.
• Own architecture and technical decisions, balancing quality, cost, and latency across the stack.

Usersnap builds AI-driven capabilities into its product, from automated setup to intelligent surveys and self-generating reporting. It is part of saas.group, a fully remote portfolio of 25 B2B SaaS brands with nearly 380 colleagues in 50+ countries and a high-trust culture.

• Build tooling for capturing and processing data from agents and humans at significant scale.
• Solve hard problems around compute, orchestration, scaling, security, and reliability.
• Help develop approaches for training, benchmarking, and evaluating AI agents.

Prolific builds human data infrastructure for AI development, connecting researchers with a global pool of participants to collect high-quality, ethically sourced behavioral data. They are a mission-driven company at the forefront of AI innovation, with a remote culture and a focus on impactful work.

• Deliver end-to-end functionality across a TypeScript/React front end and a Golang backend.
• Build products for partner integrations, including wallet systems and game systems in a certified environment.
• Design and build APIs (GraphQL, gRPC, REST) and shape the architecture of a young platform on AWS.

Oddin builds products that partners integrate into their own platforms, focusing on AI-first development where 80-90% of code is generated with AI. They are a young startup with a collaborative, autonomous product team and no strict processes.

• Architect production backend services and data models for financial workflows.
• Build full-stack features and AI agent-powered credit risk operations.
• Use modern AI-assisted development tools and contribute to engineering standards.

Gynger provides embedded financing and cash flow management for B2B tech vendors, using AI agents for credit risk. Founded in 2021 and based in NYC, the Series A startup has raised $20M and is a Built In Best Place to Work.

• Build and ship AI-enabled applications end to end using LangChain, LangGraph, and LangSmith.
• Own technical delivery across the full stack: backend services with Python/FastAPI and modern web UIs with TypeScript/React.
• Collaborate closely with designers, product managers, and Flynn stakeholders to continuously deliver working software.

Flynn is building AI agents for field service work, giving technicians the knowledge they need to fix complex equipment on-site, the first time, every time. It is a spinout from Equips, a leader in equipment maintenance services, and is rolling out to its first customers.

• Own features on Techstars OS end to end, from data ingestion to UI, including AI-native features like semantic search and assistive workflows.
• Lead the migration and shutdown of legacy founder tools, consolidating functionality into the OS or retiring cleanly.
• Act as a technical resource for the team, unblocking others and maintaining context across time zones.

Techstars builds an operating system that centralizes how accelerator programs, founder support, network engagement, and investment decisions are run. It is a small, high-trust team that moves fast and ships real things.

• Own reusable utility ML libraries for forecasting, disaggregation, demand response, detection, and asset health. - Own the simulation engine with Rust and Python bindings, including validation methodology. - Own the grid-data toolkits including protocol codecs and synthetic scenario generators.

AZX accelerates positive impact in critical industries through AI transformation, specializing in physics-informed ML for climate and sustainability. Founded in 2024, it is a profitable public benefit corporation that works with industry leaders in real estate, energy, logistics, and utilities.

• Own Merlin's foundation and world-model work, including architecture selection, post-training, and capability roadmap.
• Lead and mentor a small team of world-model engineers, setting the technical bar and review culture.
• Design model interface to the autonomy stack with structured, schema-constrained plan outputs and build evaluation harnesses.

Merlin is a publicly traded aerospace and defense company building a non-human pilot for full-stack aircraft autonomy. Headquartered in Boston, it is expanding its organization to accelerate the deployment of its autonomy platform.

• Own the infrastructure layer for AI workloads including inference serving, Kubernetes, and agent-sandboxing platforms.
• Manage the serving tier for open-weight models, Kubernetes operators, and stateful data planes.
• Oversee the sandbox runtime, control-plane services, and observability tooling.

AZX accelerates positive impact in critical industries through AI transformation, specializing in physics-informed ML and enterprise AI solutions for climate and sustainability. Founded in 2024, the company is a profitable public benefit corporation with a growing team working with category leaders in real estate, energy, logistics, and utilities.

• Lead a team of backend and full-stack engineers, driving architectural decisions for scalable APIs and microservices.
• Partner with product and other teams to ensure the platform remains reliable and fast, supporting new integrations.
• Stay hands-on with code, reviewing pull requests and unblocking engineers on complex problems.

Tradeify builds a high-performance trading ecosystem for futures and crypto markets, partnering with platforms like NinjaTrader and Tradovate. The company is rapidly scaling and serves a global trader community, with technology at its core.

• Direct the agent array on production workstreams by decomposing problems into tasks and integrating agent output into shipped software.
• Review agent-generated pull requests at volume and depth, identifying correctness, security, and accessibility defects.
• Author evaluation suites that make quality measurable using eval-driven development and own end-to-end quality within a FedRAMP-authorized environment.

Granicus provides cloud-based solutions for government communications, website design, meeting management, and records management, serving over 5,500 agencies and 300 million citizens. With a globally distributed team and a culture of transparency and inclusion, Granicus has been recognized on the GovTech 100 list for the past 5 years.

• Lead a team of AI engineers, setting technical direction and quality standards while staying hands-on in code.
• Own the architecture and end-to-end delivery of AI systems using LLMs, RAG, and agentic patterns.
• Drive evaluation, monitoring, and continuous improvement loops to ensure production-grade AI capabilities.

Finom is a European tech startup building an all-in-one financial B2B platform for entrepreneurs, integrating banking, accounting, and invoicing. They are a well-funded company with over $346 million in total funding, maintaining a startup culture that values innovation, swift implementation, and employee impact.

• Define the AI Stockbroking proposition, product roadmap and business model.
• Build AI-native experiences across stock discovery, research, trading, portfolio management and market information.
• Own P&L, funded accounts, assets, trading activity, revenue and unit economics.

BJAK is a mobile-first insurance platform that enables insurance to be accessible online, and is the leading insurance platform in Southeast Asia. The company has teams working around the world with over 20 nationalities, fostering a culture of passion and innovation.

• Design and implement LLM solutions including RAG, agents, and prompt orchestration.
• Integrate AI provider APIs (OpenAI, Anthropic, Google) into production systems with focus on cost and latency.
• Evaluate architectures and define metrics for model quality and production monitoring.

Redbee is an Argentine technology company with over 14 years of experience redefining digital products in the financial industry. They are passionate about co-creating innovative solutions and focus on delivering value quickly and with quality.

• Shape AI security at the edge by turning new threats into protections customers can use.
• Think like an attacker to find new ways AI systems can be exploited and stop them.
• Build for real time, catching threats without slowing down the applications we protect.

bunny.net is a content delivery network and edge computing company that accelerates the internet. Founded in 2015, the company has 95+ employees and fosters a culture of ownership, improvement, and challenging what's possible.

• Design end-to-end LLM evaluation plans for business scenarios such as dialogue and financial trading.
• Build evaluation metric systems and rubrics to quantify model performance and drive improvement.
• Lead construction of evaluation datasets, define dimensions, and ensure high-quality annotation standards.

Binance is a leading global blockchain ecosystem behind the world's largest cryptocurrency exchange by trading volume and registered users. Trusted by 300+ million people in 100+ countries, we offer a diverse range of digital-asset products and services.