Mid-Level Software Engineer

About EcoMetricx

EcoMetricx is a data-science and AI company that helps utilities, municipalities, and enterprises turn messy operational data into decisions they can defend. We build production software across a portfolio of internal products and custom client deliverables, including automated program-evaluation platforms, AI-powered network analytics, natural-language data exploration tools, and multi-tenant dashboards and chatbots for clients across energy, real estate, content, healthcare, and the public sector.

Our work spans causal inference, forecasting, AI/LLM integration, and the hard plumbing underneath: secure data pipelines, multi-tenant architecture, and compliant deployments. Over the next twelve months, our roadmap includes shipping a cloud-native ML evaluation pipeline, migrating a major enterprise client to a modern lakehouse platform, standing up a HIPAA-compliant messaging system for a healthcare partner, expanding into new international markets, and deepening integrations with strategic technology partners.

What You'll Work On

You'll rotate across two or three concurrent client engagements at any given time. Expect to be involved in things like:

•       Redesigning ML pipelines that power our evaluation products so they handle larger data volumes with cleaner tenant isolation.

•       Building secure, auditable CI/CD workflows (GitHub Actions, SOC 2-aligned approvals) so client code changes can ship safely without a bottleneck on one person.

•       Hardening authentication and data-handling paths. We've been tightening token handling, encryption-at-rest, and role-based access in response to client compliance needs (SOC 2, GDPR, CPUC, HIPAA-adjacent work).

•       Integrating LLM features into customer-facing products: retrieval pipelines, hallucination guardrails, token-cost optimization, and multi-tenant chatbot orchestration.

•       Owning API performance work. We solve annoying problems around SQL query shape, concurrency limits, and ingestion from third-party data providers and partner APIs.

•       Building ingestion and transformation layers that bend to evolving client logic models, not the other way around.

•       Helping us produce technically rigorous RFP responses: reading a scope, spotting the risk, and sketching the architecture before a line of code is written.

What We Are Looking For

We care more about how you think than the exact stack on your resume. That said, here's what maps well to our work.

Must Have

•       Roughly 3 to 6 years of professional software engineering experience, ideally with at least one role where you were the person responsible for a system working in production.

•       Strong working knowledge of Python (our primary language) and comfort with SQL.

•       Experience designing and deploying on a major cloud platform. We mostly live in AWS (including managed ML, object storage, serverless compute, and IAM) with some lakehouse and Google Cloud Storage work.

•       Practical experience with API design, data pipelines, and the tradeoffs between batch and streaming.

•       Familiarity with CI/CD, version control workflows (GitHub Actions), and the basics of infrastructure-as-code.

•       A serious respect for security and data privacy. You should understand why unencrypted tokens are a problem, what least-privilege means, and how to reason about PII.

We'd Love If You Also Have

•       Experience with LLM applications in production: RAG, prompt orchestration, evals, cost and latency tradeoffs, and guardrails.

•       Exposure to multi-tenant architecture (row-level security, tenant isolation, per-tenant cost attribution).

•       Comfort reading a statistical model someone else wrote and implementing it cleanly: causal inference, Bayesian shrinkage, time-series forecasting, or similar. You don't need to be a statistician, but you shouldn't run away from one.

•       Work touching regulated data: HIPAA, GDPR, SOC 2, or utility data confidentiality.

•       Any direct experience with energy, utilities, metering data, or geospatial analytics.

What We Care About More Than Any Checklist

•       You think before you type. When you get a task, your first move is to understand the problem, not open your editor. You write down assumptions and check them before they become bugs.

•       You write plans teammates can follow. A well-structured ticket, RFC, or design doc from you should make it easy for another engineer, or an AI coding agent, to do the build.

•       You spot the business question underneath the technical one. If a client asks for a dashboard, you're curious about the decision it's meant to support. This matters because our clients pay us for outcomes, not artifacts.

•       You communicate plainly with non-engineers. Our stakeholders range from utility operators to nonprofit program managers to PhDs. You need to translate in both directions without condescension.

•       You're honest about uncertainty. You say “I don't know yet, but here's how I'd find out” more often than “sure, no problem.”

•       You push back thoughtfully. If a scope feels wrong or a deadline is going to produce a bad result, you say so, and you bring a proposed alternative.

How We Work

•       Small, senior-weighted team (around 10 engineers plus data scientists and business development). You'll have real ownership from week one.

•       We use AI tooling heavily: Claude, coding agents, RFP analysis, and internal research. We expect you to use them too, and to be opinionated about when they help and when they don't.

•       We default to writing things down: design docs before builds, clear handoffs, and documentation that outlives the engineer who wrote it.

•       We take security and client trust seriously. If we break it, we lose the contract, and for some clients, much more.

What a Great First 90 Days Looks Like

•       Month 1: Ramp on one active client engagement. Ship a small but visible improvement. Read the last six months of design docs. Ask a lot of questions.

•       Month 2: Own a feature or subsystem end-to-end. You write the plan, we review it, you build it (with as much AI assistance as you find useful), and you ship it with tests and docs.

•       Month 3: Start taking point on the technical framing of new client scopes. Sit in on RFP reviews. Propose one improvement to how the team works.

Compensation & Benefits

•       Competitive base salary, calibrated to experience and location.

•       Health, dental, and vision for US employees.

•       Flexible PTO, with an expectation that you actually take it.

•       A real learning budget: books, courses, conferences, and AI tools.

How to Apply

Send a short note (no cover letter template, please) telling us:

1.    A system you built or significantly rearchitected, and what you'd do differently if you built it again.

2.    A time you pushed back on a requirement. What was wrong, what you proposed instead, and how it landed.

3.    How you currently use AI tools in your engineering work, and where you think they fall short.

Include a link to code, writing, or anything else that shows how you think.

We read every application. We respond to everyone we decide to move forward with within two weeks.