✨ AI Insights & Summary
This Software Engineer role at Skylight presents a compelling opportunity to leverage cutting-edge AI and technology to significantly improve public services for vulnerable young adults transitioning out of the foster care system. Working on the "Fostering the Future" platform, you'll be instrumental in co-designing and engineering a national-scale, AI-powered resource discovery tool. The position is ideal for engineers passionate about social impact, comfortable with a modern tech stack (Python, LangChain, FastAPI, Postgres, Next.js), and skilled in RAG systems, prompt engineering, and LLM integration. Skylight fosters a lean-agile, user-centric environment with a strong emphasis on technical excellence and mentorship, offering a chance to build a critical zero-to-one product that directly benefits thousands of young people annually.
About Skylight
Skylight is a digital consultancy using design and technology to help government agencies deliver better public services. We’re at the forefront of a civic movement to reinvent how all levels of government serve families, patients, and many others in today’s digital world. If you want to play a part in driving this critical movement forward, we’d love for you to join our growing team of public interest technologists. The work we do matters.
About the Job
At Skylight, software engineers design, implement, and maintain software systems to improve government services for the public. This software is built to meet the needs of young adults transitioning out of the foster care system. Skylight is partnering with the Administration for Children and Families (ACF) to build Fostering the Future, a platform that helps young adults aging out of foster care find the resources they need, when they need them. Co-designed with former foster youth and launching in fall 2026, the platform uses traditional resource database behavior with AI to help young adults in foster care ask the questions they don’t know to ask and access the resources they don’t know exist. You’ll work on both the user experience and the resource set that makes it valuable. You’ll engineer the platform end-to-end: a chat-like experience that pairs LLM features with a curated knowledge base of resources for foster youth. Day to day, that means designing RAG systems, agent interactions, tuning prompts, refining embeddings, and ensuring any LLM integration is accurate, safe, and genuinely useful. Our stack is Python (LangChain and FastAPI), Postgres with pgvector, Next.js, and embedding and chat models including Amazon Titan and Claude — but we care more about your experience optimizing LLM interactions with knowledge bases than about any specific tool.
What to Know Going In
You’ll join a small cross-functional team of researchers, designers, product managers, and engineers helping the roughly 20,000 young people who age out of foster care each year get the support they’re entitled to. This is a fast-moving, zero-to-one effort — greenfield development on a new public-facing federal platform, at national scale.
What You’ll Do
- Engineer a software product end-to-end — frontend and backend — as part of a cross-functional team that includes researchers, designers, product managers, and other engineers.
- Design and optimize the platform’s LLM-powered features — semantic search across a knowledge base, prompt engineering, and retrieval quality — so foster youth get accurate, relevant guidance.
- Build accessible, Section 508–compliant interfaces and the APIs and services behind them, keeping secure data exchanges reliable and performant.
- Evaluate and improve the safety and quality of LLM interactions, so the experience stays trustworthy for the young people who rely on it.
- Build secure, compliant software that meets federal security standards and defends against common vulnerabilities.
- Achieve technical excellence by advocating for and adhering to lean-agile engineering practices such as API-first design, simple design, continuous integration, version control, and automated testing.
- Mentor other engineers and share knowledge to strengthen the team’s engineering practice.
What We’re Looking For
Basic Qualifications
- Experience designing and improving RAG-based LLM systems, including semantic search.
- Experience designing multiple LLM/agent interactions to search and refine data sets.
- Experience using LLMs to aggregate, refine, and review data sets.
- Strong prompt improvement skills and understanding of effective LLM patterns.
- Ability to write clean, working, and reusable code.
- Proficiency in producing high-quality code by adhering to good practices like automated testing.
- Familiarity with building secure, compliant software and defending against common vulnerabilities.
- Foundational understanding of data engineering, including using various datastores for persistent data.
- Comprehension of software system components and their integration, considering performance, reliability, and availability.
- Understanding of user-centric development and the impact of code.
- Experience working on cross-functional teams (product, design, engineering, ops) in iterative delivery.
- A work approach that aligns with Skylight's core values.
Nice-to-Haves
- Familiarity with Skylight's stack: Python (LangChain and FastAPI), Postgres with pgvector, Next.js, and embedding/chat models (Amazon Titan, Claude).
- Experience with databases and/or cloud infrastructure.
- Familiarity with CI/CD best practices.
- Experience with infrastructure as code (e.g., Docker, Terraform).
- Experience delivering technology in government, regulated industries, or public-benefit settings.
- Proven ability to stay aligned and accountable on remote teams.
Don’t meet every qualification but think you can do the job? We’d still love to hear from you. If you’re excited about the role, apply. We consider candidates with a range of backgrounds and experiences.
Other Requirements
- Some roles require a degree or additional years of experience as a substitute due to federal contracts.
- All work must be conducted within the U.S. (excluding U.S. territories).
- Must be legally authorized to work in the U.S. now and in the future without sponsorship.
- Must be able to obtain a public trust clearance.
- Must successfully complete a company background check.
Logistics
Position Type
This is a full-time, exempt position.
Location
This is a fully remote position.
Travel
This role may require a small amount of travel to Washington, DC.
Expected Start Date
As soon as possible.
Care Package
Salary
Salaries are determined based on interview performance and seniority. Current ranges are:
- Associate Software Engineer: $90,000–$125,000
- Software Engineer I: $120,000–$140,000
- Software Engineer II: $135,000–$160,000
- Senior Software Engineer: $150,000–$185,000
- Staff Software Engineer: $170,000–$203,000
- Principal Software Engineer: $180,000–$230,000
Benefits
- Medical, dental, and vision insurance.
- Short-term and long-term disability insurance.
- Life and AD&D insurance.
- Dependent care FSA, healthcare FSA, health savings account.
- Dollar-for-dollar 401(k) match up to 10% of salary with no vesting period.
- Flexible paid-time-off (approx. 25 days/year) plus 11 federal holidays.
- Up to 12 weeks paid parental leave for new birth, adoption, or foster parents.
- Performance rewards: annual salary increase, annual performance bonus, spot bonuses, stock options.
- Business development / sales bonuses.
- Referral bonuses.
- Annual $2000 allowance for professional development.
- Annual $750 allowance for tech-related purchases.
- Annual $100 swag budget.
- Dollar-for-dollar charity donation matching (up to $500/year).
- Flexible, remote-friendly work environment.
- Empowering environment for public good.
Interview Tips
Our process typically involves a preliminary screen, skills interview, behavioral interview, and a reverse interview (2-3 weeks total). Visit our join page for process details, check our Career Pathways framework, and browse our case studies for preparation. Contact recruiting@skylight.digital for reasonable accommodations.
Compliance & EEO
- We participate in E-Verify. All work must be conducted within the U.S. You must be legally authorized to work in the U.S. without sponsorship and able to obtain a public trust clearance. A background check is required.
- Skylight is an Equal Opportunity Employer.