Senior Data Engineer — Retrieval & RAG Platform
Solvd
- Remote - US
- Remote
- Senior level
- 4+ yrs experience
- Data Engineering
- Visa: not specified
The role in 30 seconds
This role owns the retrieval workstream end to end for an enterprise AI engagement at Solvd, building the Bedrock Knowledge Base, OpenSearch Serverless vector collection, versioned retrieval API, and quality dashboard. The engineer must ship a retrieval API that returns the right passages with citations only to entitled callers, at acceptable latency and cost, and hand it over to the client to operate and version after week 10.
Is this role a fit for you?
The backgrounds this role is most likely to hire from, as RoleStack reads the posting.
Strong match
- B2B infrastructure / API / developer tools
- Data platform / ML pipeline
- B2B enterprise platform / workflow software
- Enterprise-scale
- Growth-stage
Workable stretch
- Creator / builder / self-serve tool
- Marketplace / multi-sided platform
- Regulated environment
Harder transition
- B2C consumer-facing application
- Embedded systems / hardware-adjacent
- Startup / greenfield
Industry: Software, SaaS & AI Platform. See your personal match score
What they're really screening for
Must-haves
- CORE 4+ years as a software or ML engineer, with at least one RAG or enterprise search system delivered to production and measured with an evaluation suite (not just demos)
- CORE Hands-on with Amazon Bedrock — Knowledge Bases, Retrieve and RetrieveAndGenerate APIs, embedding models
- CORE Amazon OpenSearch Serverless vector search including hybrid queries and metadata filters
- CORE Python, with experience shipping Lambda-based APIs behind API Gateway including auth, versioning, and throttling
- CORE Infrastructure-as-code with Terraform or AWS CDK and CI for deployment into a client-owned account
- CORE Practical understanding of chunking, embedding model selection, reranking, and trade-offs between relevance, latency, and cost
- PLUS Experience building operational dashboards in CloudWatch or an equivalent tool
- CORE Experience enforcing entitlement/permission-aware filtering at query time so restricted content is never returned to unentitled callers
- PLUS Experience building for contact center / Agent Assist users and integrating as an API provider
Nice-to-haves
- PLUS AWS Certified Generative AI Developer Professional or AWS Certified Machine Learning Engineer Associate
- PLUS Experience with RAG evaluation frameworks and metrics — faithfulness, context precision and recall, citation accuracy
- PLUS Experience integrating with a contact center Agent Assist product as an API provider
- PLUS Familiarity with Bedrock Guardrails and permission-aware retrieval patterns over Microsoft 365 content
Résumé keywords for this role
Terms the posting stresses. Include the ones you can honestly back up.
Tools & technologies
- Amazon Bedrock
- Knowledge Bases
- Retrieve and RetrieveAndGenerate APIs
- Amazon OpenSearch Serverless
- vector search
- hybrid queries
- Python
- Lambda
- API Gateway
- Terraform
- AWS CDK
- CloudWatch
- Bedrock Guardrails
- Microsoft 365
- RAG
Concepts & phrases
- Senior Data Engineer
- Retrieval & RAG Platform
- Bedrock Knowledge Base
- OpenSearch Serverless
- retrieval API
- hybrid retrieval
- metadata filtering
- golden question suite
- retrieval and API runbooks
- permission-aware retrieval
- Agent Assist provider
- citation payload
- quality dashboard
- entitlement filtering
- evaluation suite
Words that count as the same thing: "RAG" ≈ Retrieval-Augmented Generation, Retrieval Augmented Generation · "IaC" ≈ Infrastructure-as-Code, Infrastructure as Code · "Agent Assist" ≈ Contact center agent assist, Agent-assist product · "hybrid retrieval" ≈ vector plus keyword search, hybrid search · "entitlement filtering" ≈ permission-aware retrieval, access-control filtering
Check my résumé against these keywordsHiring at Solvd
Open senior roles RoleStack tracks right now, posted in the last 30 days.
More open roles at Solvd
Full job description from Solvd
From Solvd's posting. View the original
Solvd Inc. is a rapidly growing AI-native consulting and technology services firm delivering enterprise transformation across cloud, data, software engineering, and artificial intelligence. We work with industry-leading organizations to design, build, and operationalize technology solutions that drive measurable business outcomes.
Following the acquisition of Tooploox, a premier AI and product development company, Solvd now offers true end-to-end delivery—from strategic advisory and solution design to custom AI development and enterprise-scale implementation. Our capability centers combine deep technical expertise, proven delivery methodologies, and sector-specific knowledge to address complex business challenges quickly and effectively.
Show full description Show less
We are looking for a Senior / Staff Data Engineer to build what the Agent Assist provider calls: the Bedrock Knowledge Base configuration, the OpenSearch Serverless collection, the versioned retrieval API, and the quality dashboard. You’ll own the retrieval workstream end to end, paired with the ingestion engineer, with technical direction from the Solution Architect.
This role carries the most technical risk in the engagement — entitlement filtering is a severity-one commitment. The mission: ship a retrieval API that returns the right passages, with citations, only to callers entitled to see them, at a latency and cost the client accepts, and that the client can version and operate after week 10.
WHAT YOU’LL DO
- Stand up the Bedrock Knowledge Base and OpenSearch Serverless vector collection in the client’s account from infrastructure code, with the embedding model, chunking strategy, and metadata fields the architect specified.
- Implement hybrid retrieval — vector plus keyword — with metadata filtering by taxonomy, source, status, and review date; tune top-k, score thresholds, and reranking against the golden question suite.
- Build the retrieval API on API Gateway and Lambda per the Integration Blueprint — versioned paths, authentication for the Agent Assist provider, per-client rate limits and usage metrics, structured error responses, and a citation payload with source URL, title, section, and last modified date.
- Enforce entitlement at query time by filtering on permission metadata the ingestion pipeline captures — no restricted passage is ever returned to an unentitled caller.
- Provide the Bedrock retrieve-and-generate reference endpoint used for pilot validation in weeks 7–9.
- Build the quality dashboard — relevance, grounding, and citation scores trended per run; permission test results; freshness lag per source; API latency and error rate; OpenSearch capacity.
- Support the contact center pilot in weeks 7–9 — capture retrieval misses, tune, and re-run the suite before each change ships.
- Write the retrieval and API runbooks including model change, re-embedding, and index rebuild procedures with cost implications.
- Pair with client platform owners in weeks 9–10 and hand over on the transition readiness checklist.
REQUIRED SKILLS & EXPERIENCE
- 4+ years as a software or ML engineer, with at least one RAG or enterprise search system delivered to production and measured with an evaluation suite — not just demos.
- Hands-on with Amazon Bedrock — Knowledge Bases, Retrieve and RetrieveAndGenerate APIs, embedding models — and Amazon OpenSearch Serverless vector search including hybrid queries and metadata filters.
- Python, with experience shipping Lambda-based APIs behind API Gateway including auth, versioning, and throttling.
- Infrastructure-as-code with Terraform or AWS CDK and CI for deployment into a client-owned account.
- Practical understanding of chunking, embedding model selection, reranking, and the trade-offs between relevance, latency, and cost.
- Experience building operational dashboards in CloudWatch or an equivalent tool.
PREFERRED QUALIFICATIONS
- AWS Certified Generative AI Developer Professional or AWS Certified Machine Learning Engineer Associate.
- Experience with RAG evaluation frameworks and metrics — faithfulness, context precision and recall, citation accuracy.
- Experience integrating with a contact center Agent Assist product as an API provider.
- Familiarity with Bedrock Guardrails and permission-aware retrieval patterns over Microsoft 365 content.
When you join Solvd, you'll…
- Shape real-world AI-driven projects across key industries, working with clients from startup innovation to enterprise transformation.
- Be part of a global team with equal opportunities for collaboration across continents and cultures.
- Thrive in an inclusive environment that prioritizes continuous learning, innovation, and ethical AI standards.
Ready to make an impact?
If you're excited to build things that matter, champion responsible AI, and grow with some of the industry’s sharpest minds. Apply today and let’s innovate together.
Solvd is an equal opportunity employer.
FAQ
How much experience do you need?
The posting asks for 4+ years in software or ml engineering. It is a senior-level role.
Does Solvd sponsor visas for this role?
The posting does not say whether visa sponsorship is available. Ask the recruiter before you apply.
Is this role remote?
It is a remote role.
What tools and technologies does the posting mention?
Amazon Bedrock, Knowledge Bases, Retrieve and RetrieveAndGenerate APIs, Amazon OpenSearch Serverless, vector search, hybrid queries, Python, Lambda, API Gateway, Terraform, AWS CDK, CloudWatch, Bedrock Guardrails, Microsoft 365, RAG.