Forward deployed engineer for AI, data, and production systems

I ship production systems across product, data, and infrastructure.

Forward deployed engineer for teams building AI products and data-heavy software. I scope the problem, build the system, and get it live.

See the work Start a project

Available for select projects·Async-friendly

@ashutoshbele @ash_codess LinkedIn Email

Data + infrastructure

The data and infra work behind the product

This is the part that usually breaks first if it is weak: pipelines, orchestration, cloud boundaries, observability, and migration work that has to keep running after launch.

Streaming + ingestion systemsLakehouse + modelingOrchestration + observability

View data and infrastructure work

What the work usually involves

Ingestion, orchestration, platform boundaries, monitoring, and safe modernization.

What matters most

Reliability, clear ownership, and systems teams can operate without heroics.

Sanius Health•Senior data engineer

Healthcare cloud data platform

Unified wearable event streams and scheduled operational feeds into one governed platform for analytics and product reporting.

Scope

Wearable telemetry plus operational batch sources

Reliability

Freshness gates, contract validation, and shared incident runbooks

Read the breakdown

Sanius Health•Data engineer

Clinical + wearable medallion pipeline

Built a Bronze/Silver/Gold pipeline for clinical and wearable data so sensitive analytics could move from raw capture to decision-ready models with governance built in.

Scope

Sensitive clinical and wearable datasets

Reliability

Layer freshness checks, lineage visibility, RBAC-aligned access

Read the breakdown

Selected work

Work that shipped under real constraints

A few examples of what I owned, what had to work, and what changed after it shipped.

Work inside real constraints

Own the build end to end

Ship across app, data, and infra

01•Open-source developer tools•Independent builder

CrewDock

Built an open-source macOS developer tool for persistent local workspaces, terminal sessions, and repo context.

Native macOS workspace switcher for folder-backed local development.

Workspace model

Folder-backed tabs and persistent local context instead of disposable terminal sessions

Desktop surface

Terminal panes, files, source control, tasks, and restore flows kept in one native workspace

Project status

Already usable as a real local developer tool, with the workflow still being refined in public

Open sourcemacOSDeveloper tools

View case study View source

02•Consumer health•Founding engineer

Revage

Shipped an AI health product that turned wearable data, behavior signals, and recommendations into a daily product loop.

Longevity app that turns wearable and habit data into clear daily guidance.

User clarity

A single daily score with concrete next steps

Product loop

AI guidance tied directly to behavior and wearable context

Delivery speed

Fast iteration across UX, backend, and data layers

MobileAIWearables

View case study Visit live site

03•B2B SaaS analytics•Product engineer

Stackr

Built SaaS analytics software that made software spend, usage, and ROI clear for finance and ops teams.

SaaS analytics product for understanding spend, usage, and ROI across tools.

Visibility

Feature-level usage insights for software decisions

Business fit

Reporting framed for finance, procurement, and operations

Decision speed

Clearer signals on waste, value, and adoption

SaaSAnalyticsROI

View case study Visit live site

04•AI companion product•Founding engineer

Solace AI

Built an AI companion product with stronger trust, guardrails, and response quality in production.

AI companion experience designed around warmth, clarity, and safety.

Experience quality

Supportive flows with stronger tonal consistency

Safety posture

Guardrails built into the interaction model

Iteration loop

Faster tuning through prompt and evaluation workflows

AIProductUX

View case study Visit live site

Current experiments

Products and tools still in the workshop

Early products, internal tools, and open-source ideas that are still being tested before they deserve a deeper write-up.

Fast bets, sharp first versions, and honest iteration.

FrameForge

In progress

Open-source demo recorder that captures your screen and your narration in one take.

Records you explaining the demo while also recording your workflow in real time
Built as an open-source alternative to Loom for transparent, customizable demo sharing

In progressOpen sourceDemo recorder

View project

Netwrkly

In progress

Marketplace for matching brands and creators around campaign fit and execution needs.

Helps influencers discover brand opportunities aligned to audience and content style
Enables brands to find creators based on campaign requirements and collaboration criteria

In progressCreator economyBrand collaboration

View project

PitchPrep

In progress

AI speaking coach focused on clarity, confidence, and delivery under pressure.

Improves vocabulary, talking style, tone, and accent through guided practice
Designed to build confident speakers for interviews, presentations, and high-stakes conversations

In progressSpeaking coachAI feedback

AEO Geo Scanner

In progress

Visibility tool for improving presence in AI answer engines and local search surfaces.

Boosts ChatGPT, Claude, and Gemini answer-engine presence by generating stronger citation opportunities
Improves SEO posture with structured optimization signals and geo-aware visibility insights

In progressAEOSEO

Contact

Need someone who can step in, make sense of the mess, and ship?

I work best with small teams that need fast decisions, solid execution, and systems that keep working once they are live.

Start a conversation See selected work

Best first note: what you are building, what is stuck, and what needs to work in production.

X Instagram LinkedIn

ashutoshbele8@gmail.com