High-level, Chroma lets you incorporate your own data into AI. That can be business data, like your help docs, or customer data, such as private documents.
I think there's a misconception among many non-technical people that they should "fine tune" models - to make an LLM sound like their brand or know their internal data. But, the best practice today is "context engineering" - to use a commodity LLM, and add propriety information into the prompt.
The hard part of context engineering is knowing which data to incorporate. For example, all help docs in every context creates context rot [1] that hurts accuracy, so you need a way to find which information to put in which context. The solution is essentially an internal search engine, which is Chroma.
Chroma gives AI multiple ways to search, and you can either hard-code which one works best for you - or let an agent write its own queries to conduct research by itself. Vector search is a generation ahead of old search infrastructure, and is useful for relatedness queries - like "help docs about billing". Full-text search is useful for proper nouns, like "Next.js". Regex search is useful for code and laws. Metadata search is more nuanced, but becomes really important in document search (e.g., PDFs). Chroma lets you run all of these search methods against private data, and you can use it to even include citations in results.
So, the high-level answer is: Chroma enables you to incorporate your business or customer data into AI.
Not sure why this is voted down. The site doesn't really explain what problems of mine it'll solve (if any). I just came away thinking "oh, it's a thing. might have been a useful thing, who knows"
"Sell the sizzle, not the steak" is a real thing for a reason.
your customers might have heard of RAG (retrieval augmented generation) before. chroma powers the “R” in RAG. It enables language models to dynamically pull in information to help them answer questions and solve tasks.
Fair, but I find, and I think this thread is evident, that a lot of folks are actually proactively seeking part-time work for one reason or another. If done correctly, you can actually make more money than an equivalent full-time role.
It's more likely to work if you move to Micronesia or Mexico or something like that, easy places for citizens to move to with low cost of living so you can bid on par with the other gig workers in Pakistan/India/Romania etc.
I think trying to do gig work from the states unless it is Defense contracting you're quickly going to find someone buying that might not care what geography they're drawing from which puts USA based workers at a massive disadvantage.
Or, you take a couple contracts and make more than you did full time. This doesn’t add up. If I’m part time 25 hours a week, I can squeeze another 25 hours a week in there. 2x my rate.
I empathize. The tech market is rough, one of the worst I've seen. No one knows what's going to happen so no one wants to invest, plan, hire, etc.
There's work available if you were determined enough but it may not be in your field or it may be contract work temporarily. The struggle is real and if you're in a position where you're sinking then take anything to keep you afloat.
Indeed. For now my work is a mix of 2 part time jobs, one freelancing, and one not in my field. That should sustain me for 2025, but I am keeping my eye out for ideas while I revamp my portfolio. The disappointment is normalized at this point.
With the outsourcing I'm seeing, we already are. I feel this is and even bigger factor than AI in the terrible tech market right now, only really behind zirp.
Global software dev, off-shoring, and out-sourcing have been going strong for decades. The tech market was extraordinarily strong for most of that time.
The other part that everyone misses is that the rest of the world now has better infra infra and time to catch up. The amount of people online who can program and speak English has vastly increased.
A total of 9 open part time roles, 2 of which are C-level executives, doesn't fill me with much confidence in addressing this issue. It's pretty much for those already at top or even retired to grab some side income at best.
Even if workers were choosing more paprt time hours, it's only because they were able to make enough money in full-time employment (ususally) to build up enough money to only need part time work.
I'm personally not a huge fan of the over-employment trend, although I do see some arguments for it. A lot of the companies that are looking to hire on fractional jobs are quite explicit that they don't want folks that have full-time jobs (aka are over-employed) because of the negative connotations and consequences.
I'm familiar with Higher Fraction. I believe that the core difference is just that they're an agency that takes a percent markup on your hourly rate, so typically it's 20%.
I think the best software engineers don't need to do this, though, and they can find clients directly through their network or even through places like Fractional Jobs where we connect you directly with clients, and it's your relationship to own.
Here's the thing, and I'm open to thoughts here. But I've found that candidate A can be a great fit for company B, but a terrible fit for company C. I think a match is way more subjective than folks tend to give it credit for, and a rating system (like Uber) makes it seem more objective than it actually is. It works for things like Uber, where you can have hundreds of ratings very quickly, and so it converges on a natural truth. But with fractional clients, I think the N is so small that there's might just be too much noise.
Agree with this. It's kind of like dating. No such thing as a 4.5/5 human since it's way too subjective and the enumeration of hard skills/soft skills is too high. Sometimes a fit can be an unspoken thing. You'll probably end up wasting a ton of time trying to figure out this magical algo and likely end up nowhere with it. Just my 2c
Hot take: I find that a lot of the businesses that are a good fit for terrible staff, have terrible business practices. It would probably be beneficial for a lot of absolutely garbage clients to consider hiring better people to fix their processes and corporate culture, rather than just picking the person they like.
> a good fit for terrible staff, have terrible business practices
I think generally the people that end up choosing terrible staff that perpetuate terrible business practices are probably pretty terrible executives/founders themselves. Talent generally attracts talent in some way. There's really almost no situations where someone really fantastic would even _want_ to work for a bad company barring huge pay, or a rare moment of desperation due to life circumstances
BlockBear, Firefox Focus, Hush and Wipr on Safari mobile. And it’s still not enough, though from time to time I run into issues like this where it seems like I’m asking for it not to work. Thanks for looking in to it, I like the site.
I'm curious if anyone has any experience either as a "fractional CISO" or at a company who had one.
CISO is famously a "sacrificial lamb" sort of job, and it's certainly never one I would take on on a fractional basis (all of the risk but a lot less of the reward). I could understand having a fractional "security advisor" or "security lead", so maybe I'm just arguing about semantics, but again I'm primarily just interested in hearing about real world experiences with this sort of arrangement.
I know someone who actually enjoyed the fractional CISO role. Unless your goal is to move to a less expendable CxO role I actually see it as less of the risk and more of the reward.
I realize I should probably comment links to some of the better engineering roles we're currently featuring right now. BTW I should also note we don't take a % commission like Upwork, Toptal, etc. So if you get hired you'd work with the company directly and get paid by them direct.
As someone who's done freelancing previously for years and also has run an agency previously for years, it really depends. I've certainly seen freelancers at double or triple these rates. If you already have clients, are well known, or you're good at selling the value of your time, you can ask for much higher. If you're just getting started or you're going through an agency, these rates seem pretty competitive. Also macro economic factors will change the equation and what you can ask for.
For those just getting started, my piece of advice is to be OK taking a lower rate initially, and just keep pushing it higher until you find resistance. If you're good at what you do, you will quickly find that you will get referrals (make sure to ask!) and can charge a ton more. It's a lot easier as a freelancer/contractor than a salaried employee since the market is much more liquid (you spend less time at one gig) and therefore you can test the waters with a higher rate much more often.
Regardless, what these companies list as what they will pay hourly isn't necessarily what you have to ask for. If you think about it from a negotiation perspective (and you have the ability to sell yourself), these are simply just the lower bound of what you can ask for.
That was my thought too. General rule of thumb for independent contracting is that you should take the annual salary you would normally make and chop off the zeroes to get the hourly rate. So 150K/yr becomes $150/hr. That’s about double the yearly salary and pays for your increased costs (payroll tax, healthcare, retirement, vacation) as well as your bench time between jobs.
Additionally, the going rate for an interim CTO was about $50K/month last time I checked, which doesn’t get you 40 hrs / week, so is north of $600/hr.
Yeah but the details don't matter when you say "full time equivilant". Just health insurance alone will be far far far more than you are allowing. Then there's all the tax implications. Like, if you said 2500x I might consider beleiving you. 3000x I'd probably believe you.
Every contractor already knows all this. And those who don't learn quick.
I'm not clear what you're arguing here.
When I evaluate an hourly rate, I multiply by 2100 and ask myself if this is a reasonable salary & benefits total package.
So if my rate is $125 an hour, that comes to $263k, which is a base salary of around $200k plus healthcare, self-employment taxes and sick/vacation time, etc. Now my healthcare costs might be lower than others and I don't factor in retirement because I work primarily for startups, but again this is why each contractor calculates differently.
I wouldn't multiply it by more hours if it was insufficient, I would just raise my rate.
> My rule of thumb as a contractor is to take the hourly rate x2100 to get an equivalent full-time salary plus benefits, 401k, vacation, etc.
I thought you meant to include everything on top of salary. Reading it again after this thread, maybe you meant to this is the calculation for hourly from full time salary, and then you need to also do ("plus") a calculation for everything else.
Why? It's not obvious that an employee works 2080 hours, for example, because they take vacation (or work extra), so charging a 2100th of salary doesn't cover vacation benefit. Also, the employer will pay part of health insurance, and that isn't included in the salary number, so why would it be covered by a calculation from the salary.
Once again, this relates to total comp not base salary alone. Total comp includes salary and benefits which can include vacation time.
So it's not 1/2100th of a base salary, it's the hourly rate x2100 to get the total comp.
Do you know what your preferred total comp is? Take the hourly rate being offered, multiply by x2100, if it meets that then it's a decent offer.
It's a useful rule of thumb but it's just a rule of thumb. It's not meant to be perfect but I find it easy to understand and calculate so it's served me well in my ~20 years of contracting.
2080 hours is way over estimating the hours worked per year IMHO. The way I calculate it is something like this:
2080 hours is 8 hours a day, 5 days a week, 52 weeks per year. There is no annual leave or country-specific holidays in this calculation. I'd argue that very few workers hit 2080 hours per year, and those that do are probably minimum wage and working every hour possible just to survive.
In the UK, 25-30 days holiday is typical, plus another 9 days of public holiday, so I usually reckon on 45 weeks maximum of full-time work per year, which is 225 days or 1800 hours as an upper bound. I know in the US 10 days vacation and 10 days federal holiday and maybe a bonus couple of days for Thanksgiving and Christmas would put that around 235 or 1880 hours.
As an aside, while in the US 8 hours is normal, in the UK 7.5 is more typical (and very occasionally 7), so those 225 days are now 1687.5 hours per year. But personally, I only think in daily rate, not hourly, so I don't concern myself with that.
From the 225 nominal working days, there's also paid sickness time (maybe only 0-3 days per year for me) that I'd be paid for in a salaried job but not contracting, and other days when I might be doing work related things but can't invoice it. So, maybe I'd roll that down to 220 nominal days per year.
Then there's additional taxes. In the UK, you can either pay yourself a salary and pay both employee and employer taxes on it, or pay corporation tax on company profits and pay dividends (and the personal tax that arises). Dividends worse out slightly more tax efficient, but there's not a lot in in. On average, I'd say probably 10% more in tax compared to a salaried job. So, let's call it now 200 nominal days per year.
Then, you have to consider days you can't work - between contracts, and on occasions, I've arranged a contract for when I return from holiday and then client has asked to delay for another month to help their cashflow. It's hard to quantify this, I'd say on average I've had maybe 10-15 days per year where I could have worked, but didn't because I wasn't in contract.
There are days that you might consider free or fun as a salaried employee, such as day-time work parties, conferences (and expense accounts), all of which not only aren't free in terms of time, but also come out of your income. For instance, to attend the conference I used to go to with a previous employer would cost approx. 5 days income for tickets, flights and hotels, plus the 5 days of lost income.
You've then got all the other things that are more important in the US, but are just perks here - I don't do healthcare when contracting, because the UK's free service is good enough, but I took it as an employee when it was free and I just paid the tax liability on it. Many companies include gym memberships, snacks etc.
Other costs of contracting are accountancy fees (for me about 2 days income), insurance and indemnity (for me about half a day income), heating and electricity costs for multiple computers when remote working (it's a lot of hassle to bill it to the company just to save a bit of tax by calling it a company expense rather than paying personally), etc.
There's the risk of being out of work when you weren't expecting it - not finding a contract when the previous one ends, etc. I've even had the case where I've negotiated a contract to start when I return from holiday, and on the starting date the client asked if I could delay a month because of cashflow issues.
Taking into account all these, I'd say the 200 days should be reduced to at most 180 assuming no gaps in work, and maybe 150 assuming that you might not be working for 6 weeks.
I've actually heard several people say they assume 100 working days per year for straight salary to day-rate comparisons, assuming that while they might be able to make it work on a lower day-rate, using 100 days guarantees that it's worth doing. It's hard to get that kind of mutliplier in the UK though and 150 days is about where most contracts seem to be offered at by companies compared to salaries offered for similar roles.
150 days is also conveniently 2/3 of 225 days of maximum full time load in a salaried job, which converts to a simple 50% premium over a salaried job to cover the additional work, expenses and risks of contracting.
TLDR: add 50% to the salary for comparison, or multiply daily rate by 150.
The higher rates come through trusted referrals. If you arrive at a company via referral from someone they trust, they will usually pay a lot more than hiring random people who apply.
Thanks! Most of the clients hiring engineers tend to be non-technical, especially those hiring Fractional CTOs of course. But sometimes we'll work with CTOs to hire a contractor for their team too.
I think the devs that have the most success have the following qualities:
- They can clearly show their work history (Linkedin, Resume, Github, custom site, etc.)
- They have some baseline level of knowledge/experience in relationship management, i.e. they can talk the talk. You do have to do a bit of "selling", especially when looking for contract work.
